Clinical and pathological characteristics of EGFR mutation in operable early-stage lung adenocarcinoma

Accepted Manuscript Title: Clinical and pathological characteristics of EGFR mutation in operable early-stage lung adenocarcinoma Authors: Masaya Yotsukura, Hiroyuki Yasuda, Takao Shigenobu, Kaoru Kaseda, Kyohei Masai, Yuichiro Hayashi, Tomoyuki Hishida, Takashi Ohtsuka, Katsuhiko Naoki, Kenzo Soejima, Tomoko Betsuyaku, Hisao Asamura PII: DOI: Reference:

S0169-5002(17)30301-X http://dx.doi.org/doi:10.1016/j.lungcan.2017.04.014 LUNG 5352

To appear in:

Lung Cancer

Received date: Revised date: Accepted date:

2-3-2017 19-4-2017 20-4-2017

Please cite this article as: Yotsukura Masaya, Yasuda Hiroyuki, Shigenobu Takao, Kaseda Kaoru, Masai Kyohei, Hayashi Yuichiro, Hishida Tomoyuki, Ohtsuka Takashi, Naoki Katsuhiko, Soejima Kenzo, Betsuyaku Tomoko, Asamura Hisao.Clinical and pathological characteristics of EGFR mutation in operable early-stage lung adenocarcinoma.Lung Cancer http://dx.doi.org/10.1016/j.lungcan.2017.04.014 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Clinical and pathological characteristics of EGFR mutation in operable early-stage lung adenocarcinoma Running head: EGFR mutation in operable lung adenocarcinoma Masaya Yotsukura, MD1, Hiroyuki Yasuda MD2, Takao Shigenobu MD1, Kaoru Kaseda MD1, Kyohei Masai MD1, Yuichiro Hayashi, MD3, Tomoyuki Hishida, MD1, Takashi Ohtsuka, MD1, Katsuhiko Naoki, MD2, Kenzo Soejima MD2, Tomoko Betsuyaku, MD2, Hisao Asamura, MD1 1

Division of Thoracic Surgery, Keio University School of Medicine, 35 Shinanomachi, Shinjuku-ku,

Tokyo 160-8582, Japan 2 Department

of Pulmonology, Keio University School of Medicine, 35 Shinanomachi, Shinjuku-ku,

Tokyo 160-8582, Japan 3

Department of Pathology, Keio University Hospital, 35 Shinanomachi, Shinjuku-ku, Tokyo 160-8582,

Japan

Address correspondence to: Dr. Hiroyuki Yasuda Keio University School of Medicine, 35 Shinanomachi, Shinjuku-ku, Tokyo 160-8582, Japan E-mail: [email protected] Tel: +81 (0) 3-3353-1211 Fax: +81 (0) 3-3353-2502 Highlights 

EGFR mutation status in operable lung adenocarcinoma was evaluated.



EGFR mutation status was not a prognostic indicator after surgery.



L858R and Del-19 had different radiological and pathological features.

Abstract

1

Objectives: Over the past decade, the biological and clinical characteristics of lung cancer with epidermal growth factor receptor (EGFR) mutation have been well studied. However, most studies have focused on advanced inoperable cancer, and not on resected early-stage lung adenocarcinoma. We aimed to elucidate the differences in the clinicopathological characteristics and postoperative prognosis according to the EGFR mutation status in early-stage lung adenocarcinoma.

Materials and Methods: We retrospectively collected clinical and pathological data from 369 patients with pathological stage I or II lung adenocarcinoma who underwent complete resection. Clinicopathological characteristics and postoperative prognosis were compared depending on the EGFR mutation status, using the Chi-squared test and the log-rank test, respectively.

Results and Conclusion: Of the 369 patients, 160 (43.3%) had EGFR mutation, of which 64 (40.0%) were exon 19 deletion (Del-19) and 90 (56.3%) were exon 21 point mutation L858R. Although there was no difference in overall survival (OS) between patients with and without EGFR mutation (p = 0.086), tumors with EGFR mutation were associated with a lower consolidation to tumor ratio (CTR) (p < 0.001) and a higher incidence of a lepidic growth pattern by pathological evaluation (p < 0.001) compared to those without EGFR mutation. Among tumors with EGFR mutation, there was no difference in OS (p = 0.140) between Del-19 and L858R. Tumors with L858R were associated with a lower CTR (p = 0.046), and tended to have a higher incidence of a lepidic growth pattern by pathological evaluation (p = 0.073) compared to those with Del-19.

2

In conclusion, Although EGFR

mutation status was not a prognostic indicator after surgery in early-stage lung adenocarcinoma, L858R and Del-19 had different radiological and pathological features. Keywords: lung cancer; EGFR mutation; prognostic factor

3

1. Introduction Somatic mutations of epidermal growth factor receptor (EGFR) are found in a significant proportion of lung adenocarcinoma [1]. The correlation between EGFR mutation and sensitivity to EGFR tyrosine kinase inhibitors (EGFR-TKIs) was first reported in 2004 by three independent groups [2-4]. Since 2004, the biological and clinical characteristics of EGFR mutation have been extensively studied [2-7], and it has been determined that EGFR mutations are more prevalent in females, neversmokers, and individuals of Asian ethnicity [5, 6]. Inframe deletions around the LREA motif of exon 19 (Del-19) and point mutation in exon 21 (L858R) account for about 45% and 40% of EGFR mutations in non-small cell lung cancer (NSCLC), respectively. These mutations are reportedly the strongest predictive markers for a response to EGFR-TKIs, independent of other clinical and molecular features [8, 9]. Recently, in advanced inoperable NSCLC, EGFR Del-19 and L858R have been suggested to have different sensitivities to afatinib, a second-generation EGFR-TKI, and to cytotoxic agents [10]. This suggests that the biological and clinical characteristics of Del-19 and L858R might be different.

Regarding resected lung adenocarcinoma, some studies have reported that patients with EGFR mutations were associated with a better prognosis after surgery compared with patients without EGFR mutations, whereas other studies have reported that there is no difference [7, 11, 12]. In addition, the clinical and pathological characteristics according to the subtypes of EGFR mutation in early-stage operable lung adenocarcinomas have not been well evaluated. Therefore, the present 4

study aimed to analyze the postoperative prognosis and clarify the clinical and pathological characteristics of two major subtypes of EGFR mutation in operable lung adenocarcinoma.

2. Materials and methods

2.1. Patients, follow-up methods, and analysis of EGFR mutation

We retrospectively collected the records of 369 patients with pathological stage I or II lung adenocarcinoma who underwent curative surgical resection at our institute between August 2010 and December 2014. Complete resection was defined as a resection of all macroscopic tumor tissue and a resection margin free of tumor cells upon microscopic analysis.

Tumor recurrence during the routine post-surgical follow-up was recorded. The follow-up evaluation included a physical examination, blood analysis, chest radiography and computed tomography. Whenever symptoms or signs of recurrence were detected, further evaluation was performed. This included computed tomography (CT) of the chest and abdomen, brain magnetic resonance imaging, positron emission tomography of the neck, chest and abdomen, and bone scintigraphy. Based on these examinations, recurrence of tumor was diagnosed. The recurrence-free survival (RFS) period was defined as the duration from the date of surgical resection to the date the recurrence was detected. Follow-up was carried out till the end of June 2016.

For the 369 enrolled patients, we collected data for gender, age at operation, smoking habits, level of preoperative serum carcinoembryonic antigen (CEA), extent of pulmonary resection, 5

location of tumor [13], pathological size of tumor, consolidation/tumor ratio (CTR) of tumor on thinsection chest CT with a cutoff value of 0.5 [14-16], presence of a lepidic growth pattern, status of lymphovascular and pleural invasion, T status, N status, stage, and status of adjuvant treatment. Histological classification was determined according to the 4th edition of the World Health Organization Classification of Tumors [17].

Formalin-fixed, paraffin-embedded tissue sections of the resected tumors were used for the analysis of EGFR mutation. DNA extraction and analysis of exons 19, 21, and others were performed using the peptide nucleic acid-locked nucleic acid polymerase chain reaction clamp method by Mitsubishi Kagaku Bio-chemical Laboratories Inc. (Tokyo, Japan).

2.2. Statistics

The data are presented as means, counts and percentages where appropriate. For the 369 patients, RFS probabilities and overall survival (OS) probabilities at three and five postoperative years were estimated using the Kaplan-Meier method. The clinical and pathological factors were compared using the log-rank test as a univariate analysis. Factors with a p value of less than 0.05 in the univariate analysis were entered into a multivariate analysis, which was performed with the Cox proportional hazard model, to evaluate the hazard ratios (HR) for RFS and OS probabilities with 95% confidence intervals (CI).

6

The clinical and pathological variables were compared between patients with and without EGFR mutation and between patients with Del-19 and L858R using Pearson’s test, Fisher’s test, Student’s t test, or the Mann-Whitney U test, where appropriate. All statistical analyses were performed using SPSS 22.0 software (IBM Corporation, Armonk, NY).

The study protocol was approved by the Medical Research Ethics Committee of Keio University School of Medicine, and all experiments were conducted in accordance with the Declaration of Helsinki. The requirement for informed consent was waived by the committee as our study was a retrospective review of patient records.

3. Results

3.1. Effect of EGFR mutations on the prognosis of lung cancer patients

The baseline characteristics of the enrolled patients are summarized in Table 1. Of the 341 patients in whom the status of EGFR mutation was examined, 160 (46.9%) were positive for EGFR mutation. Of the 160 patients with EGFR mutation, 64 (40%) had Del-19 and 90 (56.3%) had L858R. The remaining 6 (1.6%) patients had G719X, with 1 case with concomitant L861Q and 1 case with concomitant T790M. Thirty-eight patients with EGFR mutation and 39 without EGFR mutation had adenocarcinoma in situ.

Of the included 369 cases, 161 (43.6%), 163 (44.2%), 15 (4.1%), and 13 (3.5%) cases had lepidic-, papillary-, acinar-, and solid-predominant subtypes, respectively. The 7

remaining 17 (4.6%) cases had variant subtypes, all of which were mucinous adenocarcinoma. Fifty cases (13.6%) had radiological multiple synchronous ground-glass opacity (GGO).

The median follow-up period for the 369 patients enrolled in the study was 36 months (range: 1-68). The Kaplan-Meier curves for RFS and OS classified according to the EGFR status are shown in Figure 1. While RFS was significantly longer in patients with EGFR mutation compared with patients without EGFR mutation (p = 0.033), there was no significant difference in OS (p = 0.086).

Table 2 shows the results of the uni- and multivariate analyses for RFS and cancer-specific OS after surgery. The cutoff value of CEA was chosen to be 3.5 ng/ml, which was the mean value among the enrolled patients. For RFS, the preoperative serum CEA level (p = 0.013), size of the tumor (p = 0.003), the presence or absence of smoking habits (p = 0.001), lepidic growth pattern in the tumor (p < 0.001), lymphatic permeation (p < 0.001), vascular invasion (p < 0.001), pleural invasion (p < 0.001), N1 metastasis (p < 0.001), adjuvant treatment (p < 0.001), and EGFR mutation (p = 0.033) were associated factors. In the multivariate analysis, the presence of a lepidic growth pattern (HR = 0.356, 95% CI = 0.140-0.903, p = 0.030), lymphatic permeation (HR = 3.012, 95% CI = 1.039-8.730, p = 0.042), and pleural invasion (HR = 3.204, 95% CI = 1.198-8.565, p = 0.020) were associated with poor RFS. The presence of EGFR mutation was not an independent indicator for recurrence after surgery in early-stage lung adenocarcinoma (HR = 0.731, 95% CI = 0.285-1.872, p = 0.514). 8

Regarding OS, sex (p = 0.022), preoperative serum CEA level (p = 0.013), the presence or absence of smoking habits (p = 0.006), a lepidic growth pattern in the tumor (p = 0.021), lymphatic permeation (p = 0.014), N1 metastasis (p = 0.008), and adjuvant treatment (p < 0.001) were associated factors. In the multivariate analysis, a higher preoperative CEA level (HR = 3.701, 95% CI = 1.284-10.671, p = 0.015) was associated with poor OS.

3.2. Clinical and pathological characteristics of tumors with and without EGFR mutation

Clinical and pathological factors according to the status of EGFR mutation are shown in Table 3. Compared with patients without EGFR mutation, patients with EGFR mutation included more females (p = 0.003) and nonsmokers (p < 0.001), and had a low preoperative serum CEA value (p = 0.024). Tumors with EGFR mutations had higher incidences of low CTR (less than 0.5) (p < 0.001), radiological multiple synchronous GGOs (p = 0.028), and a lepidic growth pattern on pathological findings (p < 0.001) compared to tumors without EGFR mutations. In addition, tumors with EGFR mutation were unlikely to show non-lepidic predominant subtype (p < 0.001), vascular invasion (p = 0.004), pleural invasion (p = 0.01), T3 disease (p = 0.036), or stage II disease (p = 0.008) compared to tumors without EGFR mutation.

3.3. Clinical and pathological characteristics of tumors harboring EGFR Del-19 and L858R

9

The Kaplan-Meier curves for RFS in the EGFR mutation subtypes Del-19 and L858R

are

shown in Figure 2. No significant difference in RFS was detected (p = 0.892). There was also no difference in OS between Del-19 and L858R (p = 0.140, Kaplan-Meier curve is not shown).

Table 4 summarizes the comparison of clinical and pathological characteristics between tumors with EGFR Del-19 and L858R. There was no difference between EGFR Del-19 and L858R for either sex, age at operation, smoking habits, lymphatic permeation, vascular invasion, pleural invasion, T status, N status, stage, incidence of radiological multiple synchronous GGO, or status of adjuvant treatment (p value for each characteristic was more than 0.05). Tumors with L858R had a higher incidence of low CTR (less than 0.5) (p = 0.046) and tended to have a lepidic predominant growth pattern (p = 0.055) as well as the incidence of a lepidic growth pattern (p = 0.073) compared to tumors with Del-19. Tumors with L858R had significantly larger diameters than those with Del-19 (p = 0.016).

3.4. Demographic, radiological, and pathological characteristics of minor types of EGFR mutation

Of the 6 cases with EGFR mutated-tumors other than L858R and Del-19, 3 were male and 3 had smoking habits. The average tumor size was 15.50±6.92 mm (range: 8-28). Three patients had a tumor with a papillary predominant subtype, and 2 had a lepidic predominant subtype. The remaining one had an acinar predominant subtype. All of the 6 cases had 10

tumors with CTR of more than 0.5 and were pathological N0. The average follow-up period was 2.03±1.19 years (range: 0.51-3.96). None of the 6 cases have developed recurrences during the follow-up period.

4. Discussion

Over the past decade, several studies have examined the association between postoperative prognosis and the status of EGFR mutation in lung cancer [5, 7, 11, 12, 18-27]. However, the prognostic impact of EGFR mutation in resected cases has not always been consistent in previous studies. Some reports have claimed that EGFR mutation was not related to the prognosis after surgery [19, 20], whereas others have demonstrated that tumors with EGFR mutation are associated with a better prognosis than wild-type tumors [5, 11, 12, 18]. Some studies have reported that the EGFR mutation was a prognostic factor in a univariate analysis, but not in a multivariate analysis, which was similar to the findings in our study [5, 11]. One reason for the disparity in the results of various studies might be the different populations of patients enrolled in each study. Some studies only included patients with stage I disease, while others included patients with stage I-IV cancers [5, 11, 18, 21]. Furthermore, the prognostic implication of EGFR mutation might be influenced by the follow-up period after surgery or the age or ethnicity of the patient [22, 27]. In our study, cancers without an EGFR mutation were less likely to have CTR < 0.5 or a lepidic growth

11

pattern and were more likely to have radiological multiple synchronous GGO, pleural invasion, T3 disease, and stage II disease than those with an EGFR mutation. These facts may be associated with the prognosis. Additional follow-up or detailed prospective studies will be required to draw definitive conclusions regarding the effect of EGFR mutation on the prognosis after surgery [28].

The presence or absence of adjuvant treatment was associated with postoperative RFS and OS in univariate analyses in our study. Both N1 disease and T1b disease were regarded as candidates for adjuvant chemotherapy at our institution, and therefore the prognostic impact of the adjuvant treatment might be related to the cancer stage.

In the present study, EGFR mutation was detected in 46.9% of the examined patients, which is similar to the values in previous studies for Japanese patients with lung adenocarcinoma [5, 7, 11, 21, 22]. The number of patients with L858R was slightly more than the number with Del-19, and L858 and Del-19 together accounted for approximately 90% of all of the EGFR mutations in our study. There was no significant difference in the prognosis after surgery between Del-19 and L858R mutation subtypes. These findings are consistent with some previous reports [5, 11, 20]. Our study is the first large-scale study to compare the prognosis and clinicopathological features between EGFR Del-19 and L858R in only stage I and II lung adenocarcinoma. Most of the previous reports that compared the characteristics of EGFR Del-19 and L858R included stage III or IV disease, which require multimodality treatment [5, 11, 20, 23, 29]. Thus, the analyses in those studies might be too complicated for an objective comparison of Del-19 and L858R. 12

Del-19 and L858R may have different biological characteristics, as shown in a report on their responsiveness to TKIs [29]. Moreover, there may also be clinical or pathological differences between these two groups. In our study, while there was no significant difference in T status or stage, the average tumor diameter was significantly larger in L858R than in Del-19. In addition, there tended to be more tumors with lepidic growth patterns in L858R than in Del-19, albeit this difference was not statistically significant. The incidence of tumors with a CTR of less than 0.5 was higher in patients with L858R than in Del-19. These data suggest that tumors with EGFR L858R tend to have large GGO lesions on CT. In fact, L858R has been shown to be associated with dominant GGO lesions in lung adenocarcinoma [10]. These findings suggest that there may be biological differences between EGFR Del-19 and L858R.

Of note, the analyses of EGFR mutation in completely resected stage I and II adenocarcinoma are usually not performed because the results have no prognostic impact and do not imply any changes in postoperative treatment. The results of our study would thus not have a significant influence on clinical practice. However, they might bring about some advances in the area of cancer research and can help us better understand the biological behavior of tumors.

Our study had several limitations. First, it was a retrospective study that included 28 patients (7.6%) in whom the EGFR mutation status was not examined. Second, the median follow-up period after surgery was 36 months, which might be insufficient to evaluate the OS. This may be why 13

there was no significant difference in OS between patients with and without EGFR mutation, although there was a significant difference in RFS. Furthermore, we identified only one significant factor that affected OS in the multivariate analysis. This might be due to the small number of cancerrelated deaths among the enrolled patients, which may be due to the relatively short follow-up duration and the low incidence of cancer recurrence in resected early-stage lung cancer.

In conclusion, patients with EGFR mutation had a better RFS than those without EGFR mutation, which was not an independent factor in a multivariate analysis. Tumors with the L858R mutation had a lower CTR and tended to have GGO-dominant lesions, although there was no difference in RFS or OS between patients with Del-19 and L858R. These data might highlight the distinctive characteristics of EGFR mutation subtypes in lung adenocarcinoma.

Funding: This study was supported by unrestricted institutional funds from Eli Lilly Japan K.K., Pfizer Japan Inc., Shionogi & Co., Ltd., Ono Pharmaceutical Co., Ltd., Tsumura & Co., Taiho Pharmaceutical Co., Ltd., Meiji Seika Pharma Co., Ltd., Astellas Pharma Inc., and Covidien Japan Inc.

References [1] N. Cancer Genome Atlas Research, Comprehensive molecular profiling of lung adenocarcinoma, Nature 511(7511) (2014) 543-50. [2] T.J. Lynch, D.W. Bell, R. Sordella, S. Gurubhagavatula, R.A. Okimoto, B.W. Brannigan, P.L. Harris, S.M. Haserlat, J.G. Supko, F.G. Haluska, D.N. Louis, D.C. Christiani, J. Settleman, D.A. Haber, Activating mutations in the epidermal growth factor receptor underlying responsiveness of non-small-cell lung cancer

14

to gefitinib, N Engl J Med 350(21) (2004) 2129-39. [3] J.G. Paez, P.A. Janne, J.C. Lee, S. Tracy, H. Greulich, S. Gabriel, P. Herman, F.J. Kaye, N. Lindeman, T.J. Boggon, K. Naoki, H. Sasaki, Y. Fujii, M.J. Eck, W.R. Sellers, B.E. Johnson, M. Meyerson, EGFR mutations in lung cancer: correlation with clinical response to gefitinib therapy, Science 304(5676) (2004) 1497-500. [4] W. Pao, V. Miller, M. Zakowski, J. Doherty, K. Politi, I. Sarkaria, B. Singh, R. Heelan, V. Rusch, L. Fulton, E. Mardis, D. Kupfer, R. Wilson, M. Kris, H. Varmus, EGF receptor gene mutations are common in lung cancers from "never smokers" and are associated with sensitivity of tumors to gefitinib and erlotinib, Proc Natl Acad Sci U S A 101(36) (2004) 13306-11. [5] T. Kosaka, Y. Yatabe, R. Onozato, H. Kuwano, T. Mitsudomi, Prognostic implication of EGFR, KRAS, and TP53 gene mutations in a large cohort of Japanese patients with surgically treated lung adenocarcinoma, J Thorac Oncol 4(1) (2009) 22-9. [6] T. Mitsudomi, T. Kosaka, H. Endoh, Y. Horio, T. Hida, S. Mori, S. Hatooka, M. Shinoda, T. Takahashi, Y. Yatabe, Mutations of the epidermal growth factor receptor gene predict prolonged survival after gefitinib treatment in patients with non-small-cell lung cancer with postoperative recurrence, J Clin Oncol 23(11) (2005) 2513-20. [7] M. Sonobe, M. Kobayashi, M. Ishikawa, R. Kikuchi, E. Nakayama, T. Takahashi, T. Menju, K. Takenaka, R. Miyahara, C.L. Huang, K. Okubo, T. Bando, H. Date, Impact of KRAS and EGFR gene mutations on recurrence and survival in patients with surgically resected lung adenocarcinomas, Ann Surg Oncol 19 Suppl 3 (2012) S347-54. [8] M. Santarpia, G. Altavilla, V. Pitini, R. Rosell, Personalized treatment of early-stage non-small-cell lung cancer: the challenging role of EGFR inhibitors, Future Oncol 11(8) (2015) 1259-74. [9] M. Santarpia, G. Altavilla, M.F. Salazar, I. Magri, G. Pettineo, S. Benecchi, R. Rosell, Tyrosine kinase inhibitors for non-small-cell lung cancer: finding patients who will be responsive, Expert Rev Respir Med 5(3) (2011) 413-24. [10] J.C. Yang, Y.L. Wu, M. Schuler, M. Sebastian, S. Popat, N. Yamamoto, C. Zhou, C.P. Hu, K. O'Byrne, J. Feng, S. Lu, Y. Huang, S.L. Geater, K.Y. Lee, C.M. Tsai, V. Gorbunova, V. Hirsh, J. Bennouna, S. Orlov, T. Mok, M. Boyer, W.C. Su, K.H. Lee, T. Kato, D. Massey, M. Shahidi, V. Zazulina, L.V. Sequist, Afatinib versus cisplatin-based chemotherapy for EGFR mutation-positive lung adenocarcinoma (LUX-Lung 3 and LUX-Lung 6): analysis of overall survival data from two randomised, phase 3 trials, Lancet Oncol 16(2) (2015) 141-51. [11] Y. Matsumura, Y. Owada, T. Yamaura, S. Muto, J. Osugi, M. Hoshino, M. Higuchi, T. Ohira, H. Suzuki, M. Gotoh, Epidermal growth factor receptor gene mutation as risk factor for recurrence in patients with surgically resected lung adenocarcinoma: a matched-pair analysis, Interact Cardiovasc Thorac Surg (2016). [12] S.P. D'Angelo, Y.Y. Janjigian, N. Ahye, G.J. Riely, J.E. Chaft, C.S. Sima, R. Shen, J. Zheng, J. Dycoco, M.G. Kris, M.F. Zakowski, M. Ladanyi, V. Rusch, C.G. Azzoli, Distinct clinical course of EGFR-mutant resected lung cancers: results of testing of 1118 surgical specimens and effects of adjuvant gefitinib and

15

erlotinib, J Thorac Oncol 7(12) (2012) 1815-22. [13] Y. Tsurugai, T. Kozuka, N. Ishizuka, M. Oguchi, Relationship between the consolidation to maximum tumor diameter ratio and outcomes following stereotactic body radiotherapy for stage I non-small-cell lung cancer, Lung Cancer 92 (2016) 47-52. [14] H. Asamura, K. Suzuki, S. Watanabe, Y. Matsuno, A. Maeshima, R. Tsuchiya, A clinicopathological study of resected subcentimeter lung cancers: a favorable prognosis for ground glass opacity lesions, Ann Thorac Surg 76(4) (2003) 1016-22. [15] K. Suzuki, T. Koike, T. Asakawa, M. Kusumoto, H. Asamura, K. Nagai, H. Tada, T. Mitsudomi, M. Tsuboi, T. Shibata, H. Fukuda, H. Kato, G. Japan Lung Cancer Surgical Study, A prospective radiological study of thin-section computed tomography to predict pathological noninvasiveness in peripheral clinical IA lung cancer (Japan Clinical Oncology Group 0201), J Thorac Oncol 6(4) (2011) 751-6. [16] K. Suzuki, J.L.C.S. Group, JCOG0201 defined "radiological early peripheral lung adenocarcinoma", J Thorac Oncol 6(8) (2011) 1452-3. [17] International Agency for Research on Cancer, W.D. Travis, E. Brambilla, A. Burke, A. Marx, A.G. Nicholson, WHO Classification of Tumours of the Lung, Pleura, Thymus and Heart, 4th Edition. ed.2015. [18] B. Izar, L. Sequist, M. Lee, A. Muzikansky, R. Heist, J. Iafrate, D. Dias-Santagata, D. Mathisen, M. Lanuti, The impact of EGFR mutation status on outcomes in patients with resected stage I non-small cell lung cancers, Ann Thorac Surg 96(3) (2013) 962-8. [19] M.H. Kim, H.S. Shim, D.R. Kang, J.Y. Jung, C.Y. Lee, D.J. Kim, J.G. Lee, M.K. Bae, H.R. Kim, S.M. Lim, E.Y. Kim, J.S. Park, K.Y. Chung, H.J. Kim, J.H. Kim, B.C. Cho, Clinical and prognostic implications of ALK and ROS1 rearrangements in never-smokers with surgically resected lung adenocarcinoma, Lung Cancer 83(3) (2014) 389-95. [20] J.L. Marks, S. Broderick, Q. Zhou, D. Chitale, A.R. Li, M.F. Zakowski, M.G. Kris, V.W. Rusch, C.G. Azzoli, V.E. Seshan, M. Ladanyi, W. Pao, Prognostic and therapeutic implications of EGFR and KRAS mutations in resected lung adenocarcinoma, J Thorac Oncol 3(2) (2008) 111-6. [21] T. Ueno, S. Toyooka, K. Suda, J. Soh, Y. Yatabe, S. Miyoshi, K. Matsuo, T. Mitsudomi, Impact of age on epidermal growth factor receptor mutation in lung cancer, Lung Cancer 78(3) (2012) 207-11. [22] T. Nishii, T. Yokose, Y. Miyagi, Y. Daigo, H. Ito, T. Isaka, K. Imai, S. Murakami, T. Kondo, H. Saito, F. Oshita, K. Yamada, S. Matsukuma, M. Tsuboi, H. Nakayama, M. Masuda, Clinicopathological features and EGFR gene mutation status in elderly patients with resected non-small-cell lung cancer, BMC Cancer 14 (2014) 610. [23] M. Ragusa, J. Vannucci, V. Ludovini, F. Bianconi, S. Treggiari, F.R. Tofanetti, A. Flacco, R. Colella, A. Sidoni, L. Crino, F. Puma, Impact of epidermal growth factor receptor and KRAS mutations on clinical outcome in resected non-small cell lung cancer patients, Am J Clin Oncol 37(4) (2014) 343-9. [24] J.H. Jeon, C.H. Kang, H.S. Kim, Y.W. Seong, I.K. Park, Y.T. Kim, Prognostic and predictive role of epidermal growth factor receptor mutation in recurrent pulmonary adenocarcinoma after curative resection, Eur J Cardiothorac Surg 47(3) (2015) 556-62. [25] D.Y. Shin, Na, II, C.H. Kim, S. Park, H. Baek, S.H. Yang, EGFR mutation and brain metastasis in

16

pulmonary adenocarcinomas, J Thorac Oncol 9(2) (2014) 195-9. [26] D.M. Jackman, V.A. Miller, L.A. Cioffredi, B.Y. Yeap, P.A. Janne, G.J. Riely, M.G. Ruiz, G. Giaccone, L.V. Sequist, B.E. Johnson, Impact of epidermal growth factor receptor and KRAS mutations on clinical outcomes in previously untreated non-small cell lung cancer patients: results of an online tumor registry of clinical trials, Clin Cancer Res 15(16) (2009) 5267-73. [27] Y.T. Kim, Y.W. Seong, Y.J. Jung, Y.K. Jeon, I.K. Park, C.H. Kang, J.H. Kim, The presence of mutations in epidermal growth factor receptor gene is not a prognostic factor for long-term outcome after surgical resection of non-small-cell lung cancer, J Thorac Oncol 8(2) (2013) 171-8. [28] R. Govindan, S.J. Mandrekar, D.E. Gerber, G.R. Oxnard, S.E. Dahlberg, J. Chaft, S. Malik, M. Mooney, J.S. Abrams, P.A. Janne, D.R. Gandara, S.S. Ramalingam, E.E. Vokes, ALCHEMIST Trials: A Golden Opportunity to Transform Outcomes in Early-Stage Non-Small Cell Lung Cancer, Clin Cancer Res 21(24) (2015) 5439-44. [29] Y. Yang, Y. Yang, X. Zhou, X. Song, M. Liu, W. He, H. Wang, C. Wu, K. Fei, G. Jiang, EGFR L858R mutation is associated with lung adenocarcinoma patients with dominant ground-glass opacity, Lung Cancer 87(3) (2015) 272-7.

Figure Caption

17

Figr-1

18

Figr-2

19

Figr-3Figure legend

Figure 1. Survival curves for (a) recurrence-free survival (RFS) and (b) overall survival (OS) according to the status of EGFR mutation. The log-rank test showed that patients with EGFR mutation had a good RFS compared to patients without EGFR mutation (p = 0.033), while there was no significant difference in OS between the two groups (p = 0.086).

Figure 2. Survival curves for recurrence-free survival (RFS) according to the EGFR mutation subtype. The log-rank test showed that there was no difference in RFS between patients with a deletion in exon 19 (Del-19) and those with a point mutation in exon 21 (L858R) (p = 0.892).

20

Table 1. Characteristics of the patient population Variables Age at operation years (mean±SD) 66.2±10.2 Sex, n (%) Male / Female 192 (52.0) / 177 (48.0) Smoking habits, n (%) Absent / Present 177 (48.0) / 192 (52.0) Preoperative CEA ng/ml (mean±SD) 3.5±4.2 Extent of pulmonary resection, n (%) Lobectomy or more 259 (70.2) Sublobar resection 110 (29.8) Tumor location, n (%) Left lung / Right lung 154 (41.7) / 215 (58.3) Tumor size, mm (mean±SD) 23.3±15.8 Radiological multiple synchronous GGO, n (%) Absent / Present 319 (86.4) / 50 (13.6) Predominant subtype Lepidic / Papillary / Acinar / 161 (43.6) / 163 (44.2) / 15 (4.1) / Solid / Variant, n (%) 13 (3.5) / 17 (4.6) Lepidic growth pattern, n (%) Absent / Present 106 (28.7) / 263 (71.3) Lymphatic permeation, n (%) Absent / Present 307 (83.2) / 62 (16.8) Vascular invasion, n (%) Absent / Present 314 (85.1) / 55 (14.9) Pleural invasion, n (%) Absent / Present 321 (87.0) / 48 (13.0) T status, n (%) Tis / T1 / T2 / T3 85 (23.0) / 186 (50.4) / 79 (21.4) / 19(5.1) N status, n (%) N0 / N1 349 (94.6) / 20 (5.4) Pathologic stage, n (%) AIS / Stage I / Stage II 85 (23.0) / 239 (64.8) / 45 (12.2) Adjuvant treatment, n (%) Absent / Present 303 (82.1) / 66 (17.9) EGFR mutation, n (%) Absent / Present / Unknown 181 (49.1) / 160 (43.4) / 28 (7.6) Type of EGFR mutation, n (%) Del-19 64 (40.0) L858R 90 (56.3) Others 6 (3.8) SD, standard deviation; CEA, carcinoembryonic antigen, AIS, adenocarcinoma in situ; EGFR, epidermal growth factor receptor 21

Del-19, deletion in exon 19; L858R, point mutation in exon 21

22

Table 2. Uni- and multivariate analyses of recurrence-free and overall survival probabilities after surgery Recurrence-free survival Overall survival Probabilities Probabilities Univ Univ (Number of Multivariate (Number of Multivariate ariat ariat patients at risk) analysis patients at risk) analysis eP eP P P valu valu Characteristi 95% 95% 3-year 5-year HR valu 3-year 5-year HR valu e e c CI CI e e 91.1 97.3 Overall 89.2 (32) 91.1 (35) (168) (184) Age at operation ≤ 65 years 91.4 (76) 89.9 (17) 95.4 (81) 88.6 (17) 0.89 98.9 0.09 > 65 years 90.9 (92) 88.3 (15) 93.2 (18) 1 (102) 7 Sex Male 88.0 (81) 85.4 (12) 95.0 (89) 89.1 (15) 0.07 0.02 Female 94.6 (87) 93.2 (20) 100 (94) 93.6 (20) 1 2 Smoking habits Absent 96.6 (89) 96.6 (18) 1 99.3 (94) 96.0 (19) 1 0.00 1.96 0.706- 0.19 0.00 3.54 0.775- 0.10 Present 86.0 (78) 82.2 (14) 95.4 (88) 86.8 (16) 1 1 5.449 7 6 7 16.237 3 Preoperative CEA 93.9 99.2 ≤ 3.5 ng/ml 91.2 (23) 1 95.7 (24) 1 (128) (138) 0.01 1.59 0.722- 0.24 0.01 3.70 1.284- 0.01 > 3.5 ng/ml 83.1 (39) 83.1 (9) 97.2 (45) 86.3 (11) 3 9 3.546 8 3 1 10.671 5 Extent of pulmonary resection Lobectomy 96.7 90.4 (93) 84.8 (10) 92.5 (13) or more (101) Sublobar 0.25 0.39 92.7 (75) 92.7 (22) 98.9 (83) 91.9 (22) resection 6 1 Tumor size 95.2 97.6 ≤ 20 mm 94.0 (20) 1 89.2 (20) (104) (112) 0.00 2.22 0.917- 0.08 0.93 > 20 mm 85.7 (64) 82.9 (12) 96.9 (71) 93.5 (15) 3 5 5.461 1 9 Radiological multiple synchronous GGO 90.9 96.6 Absent 87.5 (30) 91.8 (33) (154) (165) 23

Present Lepidic growth pattern Absent Present

93.3 (20) 87.1 (3)

0.90 2

97.1 (24) 86.1 (3)

0.57 1

79.1 (40) 71.3 (5) 96.2 (128)

1 94.7 (50) 80.4 (8) 1 < 0.35 0.140- 0.03 97.6 0.02 1.93 0.676- 0.21 96.2 (27) 0.00 95.2 (27) 6 0.903 0 (133) 1 0 5.524 9 1

Lymphatic permeation Absent Present

95.5 (155)

95.5 (30)

97.7 (165)

1

93.5 (32)

1

< 3.01 1.039- 0.04 0.01 0.51 0.181- 0.78 63.4 (13) 48.9 (2) 0.00 94.5 (18) 74.3 (3) 2 8.730 2 4 8 1.479 5 1

Vascular invasion Absent Present

94.2 (154)

93.4 (29)

97.8 (166)

1

91.5 (31)

< 1.97 0.634- 0.24 0.30 72.2 (14) 64.2 (3) 0.00 93.7 (17) 87.8 (4) 2 6.135 0 4 1

Pleural invasion Absent Present

94.6 (156)

92.3 (28)

97.8 (166)

1

93.3 (30)

< 3.20 1.198- 0.02 0.06 67.4 (13) 67.4 (4) 0.00 93.5 (17) 76.7 (5) 4 8.565 0 4 1

N status N0 N1

93.2 (162)

93.2 (31)

97.2 (174)

1

93.3 (34)

1

< 2.83 0.941- 0.06 0.00 1.37 0.230- 0.72 50.7 (6) 27.0 (1) 0.00 100 (10) 63.5 (1) 8 8.559 4 8 1 8.174 9 1

Adjuvant treatment Absent Present EGFR mutation Absent Present

95.0 (149)

95.0 (26)

97.8 (158)

1

94.8 (28)

1

< < 2.13 0.875- 0.09 3.07 0.917- 0.06 72.6 (20) 61.9 (6) 0.00 94.9 (26) 71.6 (7) 0.00 0 5.186 6 7 10.309 9 1 1

86.9 (80) 85.6 (11)

1 95.5 (88) 89.0 (14) 0.03 0.73 0.285- 0.51 0.08 94.7 (74) 92.3 (19) 99.0 (81) 92.5 (19) 3 1 1.872 4 6 24

HR, hazard ratio; CI, confidence interval CEA, carcinoembryonic antigen; AIS, adenocarcinoma in situ; EGFR, epidermal growth factor receptor

25

Table 3. Comparison of clinicopathological factors depending on the presence or absence of EGFR mutation EGFR mutation Characteristic Absent Present P value Overall 181 160 Sex, n Male 106 68 Female 75 92 0.003 Age, n years (mean±SD) 65.5±10.7 67.0±9.4 0.278 Smoking habits, n Nonsmoker 68 96 Current or former smok 113 64 <0.001 Preoperative CEA ng/ml (mean±SD) 3.6±3.9 3.3±4.8 0.024 Tumor size mm (mean±SD) 25.6±20.0 21.2±9.0 0.465 CTR <0.5 59 83 ≧0.5 122 77 <0.001 Radiological multiple synchronous GGO, n Absent 163 130 Present 18 30 0.028 Predominant subtype Lepidic 56 91 Papillary 88 65 Acinar 10 1 Solid 10 3 Variant 17 0 <0.001 Lepidic growth pattern, n Absent 84 19 Present 97 141 <0.001 Lymphatic permeation, n Absent 145 139 Present 36 21 0.11 Vascular invasion, n Absent 145 146 Present 36 14 0.004 Pleural invasion, n Absent 149 147 Present 32 13 0.01 T status, n Tis 39 38 T1 86 87 T2 40 32 26

T3 N status, n N0 N1 Pathologic stage, n AIS Stage I Stage II Adjuvant treatment, n Absent Present SD, standard deviation CEA, carcinoembryonic antigen CTR, consolidation/tumor ratio AIS, adenocarcinoma in situ

16

3

0.036

169 12

152 8

0.646

39 109 33

38 111 11

0.008

147 34

132 28

0.759

27

Table 4. Comparison of clinicopathological factors according to the subtype of EGFR mutation Characteristic Del-19 L858R P value Overall 64 90 Sex, n Male 32 33 Female 32 57 0.099 Age, n years (mean±SD) 65.3±10.1 68.2±8.8 0.093 Smoking habits, n Nonsmoker 38 55 Current or former smok 26 35 0.828 Preoperative CEA ng/ml (mean±SD) 2.5±1.5 3.8±6.2 0.437 Tumor size mm (mean±SD) 19.8±8.9 22.3±9.0 0.016 CTR <0.5 28 54 ≧0.5 36 36 0.046 Radiological multiple synchronous GGO, n Absent 56 69 Present 8 21 0.099 Predominant subtype Lepidic 30 59 Papillary 33 29 Solid 1 2 0.055 Lepidic growth pattern, n Absent 11 7 Present 53 83 0.073 Lymphatic permeation, n Absent 55 78 Present 9 12 0.897 Vascular invasion, n Absent 59 81 Present 5 9 0.642 Pleural invasion, n Absent 59 83 Present 5 7 0.994 T status, n Tis 13 25 T1 37 45 T2 12 19 T3 2 1 0.541 N status, n N0 62 84 28

N1 2 Pathologic stage, n AIS 13 Stage I 47 Stage II 4 Adjuvant treatment, n Absent 55 Present 9 SD, standard deviation CEA, carcinoembryonic antigen; AIS, adenocarcinoma in situ

29

6

0.329

25 58 7

0.494

71 19

0.264