Factors Associated with Postoperative Recurrence in Stage I to IIIA Non–Small Cell Lung Cancer with Epidermal Growth Factor Receptor Mutation: Analysis of Korean National Population Data

Article information

Cancer Res Treat. 2025;57(1):83-94

Publication date (electronic) : 2024 July 10

doi : https://doi.org/10.4143/crt.2024.073

Kyu Yean Kim¹, Ho Cheol Kim², Tae Jung Kim³, Hong Kwan Kim⁴, Mi Hyung Moon⁵, Kyongmin Sarah Beck⁶, Yang Gun Suh⁷, Chang Hoon Song⁸, Jin Seok Ahn⁹, Jeong Eun Lee¹⁰, Jae Hyun Jeon¹¹, Chi Young Jung¹², Jeong Su Cho¹³, Yoo Duk Choi¹⁴, Seung Sik Hwang¹⁵, Chang Min Choi², Seung Hun Jang¹⁶, Jeong Uk Lim^,17, Korean Association for Lung Cancer, Korea Central Cancer Registry

¹Division of Pulmonary and Critical Care Medicine, Department of Internal Medicine, Uijeongbu St. Mary's Hospital, College of Medicine, The Catholic University of Korea, Uijeongbu, Korea

²Department of Pulmonary and Critical Care Medicine, Asan Medical Center, University of Ulsan College of Medicine, Seoul, Korea

³Department of Hospital Pathology, Yeouido St. Mary’s hospital, College of Medicine, The Catholic University of Korea, Seoul, Korea

⁴Department of Thoracic and Cardiovascular Surgery, Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, Korea

⁵Department of Thoracic and Cardiovascular Surgery, Seoul St. Mary’s Hospital, College of Medicine, The Catholic University of Korea, Seoul, Korea

⁶Department of Radiology, Seoul St. Mary’s Hospital, College of Medicine, The Catholic University of Korea, Seoul, Korea

⁷Proton Therapy Center, Research Institute and Hospital, National Cancer Center, Goyang, Korea

⁸Department of Radiation Oncology, Seoul National University Bundang Hospital, Seoul National University College of Medicine, Seongnam, Korea

⁹Department of Hematology-Oncology, Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, Korea

¹⁰Division of Pulmonology, Chungnam National University College of Medicine, Daejeon, Korea

¹¹Department of Thoracic and Cardiovascular Surgery, Seoul National University Bundang Hospital, Seoul National University College of Medicine, Seongnam, Korea

¹²Department of Pulmonary, Daegu Catholic University Medical Center, Daegu Catholic University School of Medicine, Daegu, Korea

¹³Department of Thoracic and Cardiovascular Surgery, Pusan National University Hospital, Busan, Korea

¹⁴Department of Pathology, Chonnam National University Medical School, Gwangju, Korea

¹⁵Department of Public Health Science, Graduate School of Public Healthy, Seoul National University, Seoul, Korea

¹⁶Division of Pulmonary, Allergy and Critical Care Medicine, Department of Medicine, Hallym University Sacred Heart Hospital, Anyang, Korea

¹⁷Division of Pulmonary and Critical Care Medicine, Department of Internal Medicine, Yeouido St. Mary’s Hospital, College of Medicine, The Catholic University of Korea, Seoul, Korea

Correspondence: Jeong Uk Lim, Division of Pulmonary and Critical Care Medicine, Department of Internal Medicine, Yeouido St. Mary’s Hospital, College of Medicine, The Catholic University of Korea, 63 ro-10, Yeongdeungpo-gu, Seoul 07345, Korea Tel: 82-2-3779-1035 Fax: 82-2-780-3132 E-mail: cracovian@catholic.ac.kr

Received 2024 January 19; Accepted 2024 July 8.

Abstract

Introduction

The treatment of choice in stage I to IIIA lung cancer is surgery. Despite curative resection, the 5 years postoperative recurrence-free survival (RFS) rates were 87.8% for patients with stage I, 54.7% for those with stage II, and 33.4% for those with stage III in non–small cell lung cancer (NSCLC) [1]. After curative resection in patients with stage II to IIIA NSCLC, adjuvant cisplatin-based chemotherapy is recommended [2]. Also, among patients with stage IB NSCLC, adjuvant systemic therapy is recommended in high-risk groups. According to the recent study in stage IB (American Joint Commission on Cancer [AJCC] 8th) NSCLC patients, adjuvant chemotherapy is beneficial in patients with high-risk factors, reducing recurrence rate and risk of mortality (hazards ratio, 0.408 and 0.176, respectively) [3]. According to the National Comprehensive Cancer Network (NCCN) guidelines, high-risk groups include patients with poorly differentiated tumors, vascular invasion, wedge resection, visceral pleural involvement, and unknown lymph node status.

Recent advances in perioperative treatment of resectable NSCLC have changed the landscape of early lung cancer management [4]. The ADAURA trial has shown significant efficacy of adjuvant osimertinib treatment in resectable NSCLC patients [5]. Among patients with resected, epidermal growth factor receptor (EGFR)–mutant, stage II to IIIA NSCLC, the 5-year overall survival (OS) was 85% in the osimertinib group, while only 73% in the placebo group [5]. It is undeniable that adjuvant treatment is effective in decreasing the chance of postoperative relapse and improving postoperative survival in EGFR-mutant patients who underwent complete resection for stage I to IIIA NSCLC.

However, osimertinib treatment for adjuvant purposes is not often covered by insurance or reimbursed in many countries, and financial burden is significant for most patients considering that there is no consensus regarding when to discontinue the adjuvant treatment [6]. Furthermore, both the prevalence and clinical characteristics of EGFR-mutant NSCLC patients vary according to region and race [7-9]. More clinical data are required to show which EGFR-mutant patient subgroups are at high risk of postoperative relapse and would benefit from adjuvant EGFR tyrosine kinase inhibitor (TKI) treatment.

This study evaluated risk factors for postoperative recurrence among patients who underwent complete resection using national Korean cancer data, with a focus among EGFR-mutant patients.

Materials and Methods

1. Patient selection

Data were obtained from the Korean Association for Lung Cancer Registry, a database created using a retrospective sampling survey by the Korean Central Cancer Registry (KCCR) and the Lung Cancer Registration Committee [10]. During the 2014-2017 period, the KCCR registered patients newly diagnosed with NSCLC. Among a total of 9,860 NSCLC patients, 3,364 underwent curative resection. Among them, 188 patients who were diagnosed as clinical stage IV or unknown stage were excluded. Finally, a total of 3,176 patients was enrolled in this study (Fig. 1). All patients were confirmed as having undergone operation for curative resection. Patients were staged according to the AJCC 7th edition TNM classification. Registered patients were followed for approximately 5 years after operation. All data in this study were obtained from the registered database with no additional review of individual data.

Fig. 1.

Study flow diagram. NSCLC, non–small cell lung cancer.

Results

2. Clinical characteristics of EGFR-mutant patients

The 867 patients with EGFR-mutant NSCLC who underwent curative surgery were analyzed (Table 1). Female sex and never-smokers were more prevalent than male sex and ever-smokers, respectively. Of these patients, 75.2%, 11.2%, and 11.8% were classified as stage I, stage II, and stage III, respectively. Among them, 94.3% were adenocarcinoma, accounting for the largest proportion. Among the EGFR-mutant subgroup, 44 (5.1%) patients showed early recurrence and 121 (14.0%) patients showed late recurrence. The median RFS was 33.02 months.

Table 1.

Baseline characteristics of EGFR-mutant patients (n=867)

3. Comparison between early, late, and no recurrence groups among EGFR-mutant patients

There was no significant difference in age, sex, ECOG performance status, or pulmonary function among the early recurrence group (ERG), late recurrence group (LRG), or no recurrence group (NRG) (Table 2). In the ERG, the proportion of patients with smoking history was higher than in the LRG and NRG (p=0.015). Also, the proportion of patients with higher T, N, and TNM stage cancer was higher in the ERG than the other groups (p < 0.001). Regarding pleural invasion, 15.9% of patients in the ERG were PL2 or PL3, which is significantly higher than the invasion seen in the LRG or NRG. The proportion of patients with perioperative chemotherapy and radiotherapy (RT) was higher in the ERG compared to the LRG and NRG.

Table 2.

Clinical characteristics of ERG, LRG, and NRG among EGFR-mutant patients

6. Clinicopathologic risk factors associated with postoperative recurrence in EGFR-mutant NSCLC patients after curative surgery

Table 3 shows univariate and multivariate analyses of clinicopathologic risk factors for recurrence. In multivariate analysis, age, pathologic N and TNM stages, pleural invasion, and surgery type were significantly associated with RFS. The HR of recurrence in patients with pathologic N2 category was 2.53 compared to patients with pathologic N0 (p=0.019). Also, the HR of recurrence in patients with pathologic stage II was 1.93 compared to patients with pathologic stage I (p=0.021). Patients with VPI had a 1.89-fold higher risk of recurrence compared to patients without pleural invasion (p < 0.001). Also, the HR of recurrence in patients with pneumonectomy was 0.45 compared to patients with wedge resection (p=0.006).

Table 3.

Univariate and multivariate analyses of risk factors associated with recurrence (n=867) in EGFR-mutated NSCLC patients after curative surgery

7. Clinicopathologic risk factors associated with mortality in EGFR-mutant NSCLC patients after curative surgery

Table 4 shows univariate and multivariate analyses of clinicopathologic risk factors for mortality. In multivariate analysis, sex, age, BMI, ECOG performance status, DLco, pathologic TNM stage, and pleural invasion were significantly associated with OS.

Table 4.

Univariate and multivariate analyses of risk factors associated with overall survival (n=867) in EGFR-mutated NSCLC patients after curative surgery

Female sex or patients with higher BMI or DLco had a lower risk of mortality in EGFR-mutant NSCLC patients after curative surgery. Patients with older age, higher ECOG performance status, higher pathologic stage, and PL3 had a higher risk of mortality.

Discussion

The present study evaluated a nationwide lung cancer database to assess patients who are at risk of postoperative relapse, specifically for EGFR-mutant populations. We showed that, among various clinical parameters, patients with older age, higher N and pathologic stages, and those with presence of VPI are more likely to experience postoperative relapse and are in need of adjuvant EGFR TKI treatment. In regard to OS in EGFR-mutant NSCLC patients after curative surgery, male sex, older age, low BMI, high ECOG performance status, low DLco, high pathologic stage, and pleural invasion were significant prognostic markers.

There are some notable studies regarding the assessment of risk factors associated with postoperative relapse specifically for EGFR-mutant NSCLC patients who underwent resection. In a retrospective analysis of 531 patients, N category along with lympho-vascular invasion and cytokeratin 5/6 positivity were independent predictors of locoregional recurrence [12]. In another retrospective study, smoking history, vessel invasion, and lymph node metastasis were associated with RFS in EGFR-mutant patients [13]. In addition, Yu et al. [14] indicated that smoking history, large tumor size, elevated lymph node ratio, and platelet to lymphocyte ratio in stage pIIIA-N2 NSCLC patients after complete resection are associated with early recurrence. In the present study, among curatively resected EGFR-mutant NSCLC patients, the 1-year RFS was 94.9%, the 2-year RFS was 85.8%, and the 3-year RFS was 81.7%. At 24 months, 85.7% of patients with parietal pleural invasion were recurrence-free compared to 91.5% in patients with VPI and 96.2% in patients without pleural invasion. VPI is an important prognostic factor regardless of tumor size or N category [15]. In a previous study, exfoliated tumor cells were found to drain through the pleural lymphatics, which supports the association between VPI and extensive N2 involvement [16]. The visceral pleura is rich in lymphatics and can easily involve the hilar lymph nodes. As a consequence, VPI is associated with locoregional recurrence and distant metastasis [17].

EGFR TKI treatment has provided better clinical outcomes in EGFR-mutant NSCLC patients [18,19]. Osimertinib treatment has shown longer progression-free survival compared to earlier-generation EGFR TKIs [18]. According to a recent study, a significant OS benefit was shown in EGFR-mutant NSCLC patients with adjuvant osimertinib treatment after complete resection [5]. Therefore, an updated version of the NCCN guidelines suggests adjuvant osimertinib treatment in EGFR-mutant patients in stage IIB-IIIA and stage IB-IIA with high-risk factors.

However, in clinical practice, we have encountered medical insurance problems. In many countries including Korea, the nationwide medical reimbursement system does not cover the cost of adjuvant osimertinib treatment in EGFR-mutant NSCLC patients, and the recommended 3 years of prescription often entail significant financial burden for the patients. A previous study showed that certain features associated with postoperative relapse do not translate directly to need for adjuvant EGFR TKI treatment. Patients enrolled in the ADAURA study were stage IB to IIIA (classified according to the AJCC 7th edition) EGFR-mutant (Ex19del or L858R) NSCLC [20]. Specifically for the EGFR-mutant subgroup, certain enrollment criteria between ADAURA and present study do not match. In our study, there were 4.0% squamous lung cancer patients included in the EGFR-mutant subgroup, and it is likely that EGFR mutations other than Ex19del and L858R could have been included. However, it was interesting to see that enrollment criteria of the ADUARA study mostly overlapped with factors associated with postoperative relapse in our study, suggesting a possibility that patients with stage N1-2 or pleural invasion could have benefited from adjuvant EGFR TKI treatment. Clearly, more studies to better predict postoperative relapse are necessary to enable selection of patients for adjuvant EGFR TKI treatment.

When EGFR-mutant patients were stratified according to timing of postoperative relapse (early, late, or no recurrence), several features were distinguished. TNM stage, pleural invasion level, and smoking history showed significant difference between the groups. Early recurrence patients show a higher proportion of PL2 and PL3 compared to late recurrence patients. Furthermore, VPI was an independent factor associated with RFS. However, according to logistic regression analysis for early recurrence in EGFR-mutant NSCLC patients in the present study, only pathologic T, N stage and smoking history were significantly associated with early postoperative recurrence. While these risk factors need to be validated in larger population study, patients with regarding risk factors should undergo close surveillance for early postoperative recurrence.

In a previous retrospective analysis of NSCLC patients who underwent complete resection, association between VPI and EGFR mutation was shown, with a higher frequency of EGFR mutation patients among the subgroup [21]. In our study, risk of postoperative relapse was stratified by level of VPI. Considering that the presence of sole pleural invasion (PL1) is staged as T2, we should concurrently approach T stage and presence of VPI [22], and it is likely that clinicians should consider the presence and level of pleural invasion as an important factor when initiating and maintaining adjuvant EGFR TKI treatment.

In our analysis of stage IA lung cancer patients without pleural invasion, we observed that epidemiological factors, such as sex and age, might correlate with recurrence risk. ERG group showed significantly higher proportion of old-age patients, and male sex was associated with higher risk of recurrence in the multivariate analysis. In a retrospective study of NSCLC patients who underwent surgery for stage I and II cancer, females demonstrated significantly higher 5-year OS rates (76.2%) compared to males (57.3%), with gender identified as a favorable prognostic factor in multivariate analysis [23]. Among EGFR-mutant stage IA NSCLC patients, patients with lobectomy had lower risk of recurrence, compared to patients with wedge resection. This finding is consistent with the previous study, which compared locoregional RFS between wedge resection versus lobectomy for early-stage NSCLC. In the study, overall locoregional RFS was worse in wedge resection patients vs. lobectomy patients, 82.0% vs 93.4% [24]. In addition, one meta-analysis of comparison between wedge resection and lobectomy for early-stage NSCLC showed that patients with lobectomy had higher OS than patients with wedge resection [25]. So, even in early-stage NSCLC, lobectomy should be actively considered in patients without contraindication such as low lung function or poor general condition.

In our study, a subgroup analysis of EGFR-mutant stage IB (AJCC 7th edition) patients revealed that the hazard ratio for recurrence in patients with VPI was 2.75 times higher compared to those without pleural invasion. Despite differences in TNM editions, this finding aligns with a recent article by Choi et al. [3], which also identified VPI as a high-risk factor for postoperative recurrence in the stage IB patients. Another notable finding is that the ERG had a higher percentage of adjuvant chemotherapy and RT compared to late-recurrence or no-recurrence groups. It is more reasonable to consider that patients with more risk factors underwent postoperative treatment at the time of adjuvant therapy, rather than attributing recurrence risk to the adjuvant treatment itself. Therefore, we believe that efficacy of adjuvant treatment is difficult to be analyzed due to the limitations of the retrospective design. Future studies are needed to evaluate the efficacy of adjuvant treatment in patients with EGFR mutations.

Among the factors associated with postoperative outcomes in the EGFR subgroup of our study, DLco was independently associated with postoperative survival but not with RFS. DLco, along with other pulmonary function parameters, has long been shown to affect postoperative course of patients who undergo lung resection [26], and our results reconfirm this finding [27]. Low DLco values are associated with several lung conditions including emphysema or pulmonary fibrosis [28]. Patients with relatively low DLco values may have underlying emphysema or other chronic lung diseases that could impact their survival. Additionally, a study found that a low preoperative DLco is predictive of postoperative cardiopulmonary complications, mortality, and poor long-term survival in surgical patients [29]. Clinicians should pay more attention to patients with poor lung function when managing lung cancer patients who undergo lobectomy.

In the present study, duration of OS was significantly shorter for EGFR-wild type patients than EGFR-mutant patients. However, the RFS showed no significant difference between EGFR-wild type and mutant patients. Further studies should be done about treatment after recurrence.

There are notable limitations of this study. First, a large proportion of overall study patients did not undergo EGFR mutation testing. The database includes patients diagnosed with lung cancer from years 2014 to 2017, during the early phase of which mutation testing was not routinely performed. Despite this limitation, a large number of EGFR-mutant patients added to the importance of the present study. Second, no detailed results on EGFR mutation subtype or next-generation sequencing results were available for our study, which deterred us from confirming the association between mutation subtype and postoperative outcomes. Third, TNM staging was based on the AJCC 7th edition TNM staging, and it is likely that some T2 patients would have been classified as T3 according to the 8th edition [30]. Lastly, in our study, there was no statistically significant difference in recurrence risk between stage I and stage III EGFR-mutant patients. Although other clinical outcomes appeared relatively more favorable for stage I patients compared to stage III patients, we believe this discrepancy may be due to the smaller sample size of stage III patients. A larger sample size in future studies could potentially address this discrepancy.

Among the population that underwent complete resection for early NSCLC with EGFR mutation, patients with advanced stage, pleural invasion, or limited resection are more likely to show postoperative relapse.

Electronic Supplementary Material

Article information Continued