Significance of Spread Through Air Spaces in Resected Pathological Stage I Lung Adenocarcinoma

Transcription

1 Accepted Manuscript Significance of Spread Through Air Spaces in Resected Pathological Stage I Lung Adenocarcinoma Gouji Toyokawa, PhD, Yuichi Yamada, PhD, Tetsuzo Tagawa, PhD, Yuka Kozuma, MD, Taichi Matsubara, MD, Naoki Haratake, MD, Shinkichi Takamori, MD, Takaki Akamine, MD, Yoshinao Oda, PhD, Yoshihiko Maehara, PhD PII: DOI: Reference: ATS S (18) /j.athoracsur To appear in: The Annals of Thoracic Surgery Received Date: 8 July 2017 Revised Date: 5 January 2018 Accepted Date: 15 January 2018 Please cite this article as: Toyokawa G, Yamada Y, Tagawa T, Kozuma Y, Matsubara T, Haratake N, Takamori S, Akamine T, Oda Y, Maehara Y, Significance of Spread Through Air Spaces in Resected Pathological Stage I Lung Adenocarcinoma, The Annals of Thoracic Surgery (2018), doi: / j.athoracsur 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.

2 Significance of Spread Through Air Spaces in Resected Pathological Stage I Lung Adenocarcinoma Running Head: STAS in stage I adenocarcinoma Gouji Toyokawa, PhD 1, Yuichi Yamada, PhD 2, Tetsuzo Tagawa, PhD 1, Yuka Kozuma, MD 1, Taichi Matsubara, MD 1, Naoki Haratake, MD 1, Shinkichi Takamori, MD 1, Takaki Akamine, MD 1, Yoshinao Oda, PhD 2, Yoshihiko Maehara, PhD 1 1 Department of Surgery and Science, Graduate School of Medical Sciences, Kyushu University, Maidashi, Higashi-ku, Fukuoka , Japan 2 Department of Anatomic Pathology, Graduate School of Medical Sciences, Kyushu University, Maidashi, Higashi-ku, Fukuoka , Japan Keywords: Lung adenocarcinoma; spread through air spaces; surgery; prognosis. Word Count: 4751 words Corresponding author: Gouji Toyokawa, MD, PhD Department of Surgery and Science, Graduate School of Medical Sciences, Kyushu University, Maidashi, Higashi-ku, Fukuoka , Japan gouji104kawa@gmail.com 1

3 Abstract Background: Spread through air spaces (STAS) is a recently described invasive pattern of lung cancer, which spreads within air spaces beyond the edge of the main tumor. In the current study, we investigated the significance of STAS in patients with pathological stage I adenocarcinoma. Methods: STAS was assessed in a total of 276 patients with resected pathological stage I adenocarcinoma. STAS was classified as either no STAS, low STAS (1-4 single cells or clusters of STAS), or high STAS ( 5 single cells or clusters of STAS) using a 20x objective and a 10x ocular lens. We evaluated the association between STAS and the clinicopathological characteristics and postoperative survivals. Results: Among 276 patients, 123 (44.6%), 48 (17.4%) and 105 (38.0%) were classified as having no, low and high STAS, respectively. The positivity for STAS was significantly associated with a larger radiological tumor diameter (P=0.008), a higher consolidation/tumor ratio (P<0.001), a higher maximum standard uptake value (P<0.001), a pathologically larger tumor size (P=0.004), the presence of pleural invasion (P=0.027) and a histologically invasive type (P<0.001), while STAS was not significantly associated with epidermal growth factor receptor mutations or programmed death ligand-1 expression (P=0.129 and P=0.872, respectively). Patients with the STAS had significantly shorter recurrence-free and overall survivals than those without (P<0.001 and P=0.002, respectively). According to a multivariate analysis, positivity for STAS remained an independent prognostic parameter for both the recurrence-free and overall survivals. Conclusions: STAS was associated with clinicopathologically invasive features and was predictive of a worse survival. 2

4 Lung cancer is the leading cause of cancer-related death in many countries, and its prognosis still remains unsatisfactory. Although molecular-targeted therapy and immunotherapy have greatly improved the prognosis of patients with advanced lung cancer, only surgical resection can provide a cure for lung cancer; specifically, patients with pathological stage I non-small cell lung cancer (NSCLC) can be curatively treated by lobar resection combined with systematic nodal dissection (1). In addition to lobectomy, wedge resection is sometimes performed for a small nodule presenting as ground-glass opacity (GGO) mixed with a small extent of solid component, which means radiologically noninvasive cancer (2). However, despite the optimal surgical resection, some patients with early-stage lung cancer inevitably encounter recurrence and eventual death; the 5-year overall survival (OS) in patients with pathological stage IA1 is approximately 90% (3). Some prognostic indicators that negatively affect the prognosis after surgical resection have been identified, including radiological invasiveness on computed tomography (CT) (2), a high metabolic activity measured by positron emission tomography (PET) (4), and pathological invasiveness (5). Further prognostic indicators are sought to stratify patients with early-stage NSCLC more accurately and to determine optimum surgical procedures, such as lobectomy or limited resection, and adjuvant therapy. Spread through air spaces (STAS) is a recently described invasive pattern of lung cancer defined as follows: STAS consists of micropapillary clusters, solid nests or single cells beyond the edge of the tumor into air spaces in the surrounding lung parenchyma (6). Importantly, STAS differs from conventional invasiveness, such as the presence of non-lepidic patterns, and the infiltration of stroma and blood vessels or structures such as the visceral pleura (7). Although the concept of STAS was introduced in 2015, similar concepts, including aerogenous spread with floating cancer cell clusters and tumor islands, have been reported (8-10). Regarding the possible cause of STAS, whether it is caused by an in vivo effect or is just an artifact of cutting through a tumor with a knife remains controversial (11, 12). Clinicopathologically, STAS was reported to be associated with male gender, a history of smoking, the presence of lymphovascular invasion, and more invasive subtypes such as micropapillary and solid patterns in patients with resected lung adenocarcinoma (13-15). Furthermore, an unfavorable prognostic impact was reported in both adenocarcinoma and squamous cell carcinoma (13-17). 3

5 Herein, we evaluated the relationship between STAS and the clinicopathological characteristics of patients with pathological stage I lung adenocarcinoma and its prognostic influence on their postoperative survival. Patients and Methods Study Patients Among 417 patients with lung adenocarcinoma who underwent surgery at the Department of Surgery and Science, Graduate School of Medical Sciences, Kyushu University between January 2003 and December 2012, 276 with pathological stage I adenocarcinoma were included in this study. Tissue specimens from the enrolled patients were retrieved from the registry of the Department of Anatomic Pathology, Graduate School of Medical Sciences, Kyushu University. The clinicopathological features, including age at surgery, gender, smoking history, carcinoembryonic antigen (CEA), pathologic tumor-node-metastasis (TNM) stage (seventh edition of the lung cancer staging system), pleural or lymphovascular invasion, histological subtype (World Health Organization Classification 2015), surgical procedure, EGFR mutation status, and programmed cell death-1 ligand (PD-L1), were examined. Local and distant recurrences were defined as described previously (18): local recurrence was defined as disease recurrence at the surgical margin, ipsilateral hemithorax, or mediastinum; distant metastasis was defined as disease recurrence in the contralateral lung or outside the hemithorax and mediastinum. The diameter of consolidation in each tumor (C) and the diameter of the whole tumor including GGO (T) were measured manually with axial 2-dimensional CT data at a 2 mm slice section; the C/T ratio was then calculated. We set the cut-off value based on the report by Suzuki et al., which used a cut-off value of 0.26 to discriminate radiological invasiveness from noninvasiveness (2). The maximum standardized uptake value (SUVmax) of the tumor was evaluated using 18 F- fluorodeoxyglucose-pet/ct scanning. The EGFR status was determined in tumor tissue using the peptide nucleic acid-locked nucleic acid polymerase chain reaction clamp method (Mitsubishi Chemical Medience, Tokyo, Japan) in 173 specimens (19). Immunohistochemistry for the PD-L1 expression was performed using an 4

6 anti-human PD-L1 rabbit monoclonal antibody (clone SP142, dilution 1:100; Spring Bioscience, Ventana, Tucson, AZ, USA) and evaluated as described previously (20). Briefly, systemic dissection of the hilar and mediastinal lymph nodes was performed at the same time as pulmonary lobectomy. After surgery, routine check-ups, including a physical examination, blood tests (including serum tumor markers), and chest X-ray, were performed at three-month intervals for the first three years and at six-month intervals thereafter. CT was performed twice a year for the first three years and then at least annually thereafter. Clinical information and follow-up data were obtained from the medical records. This study was approved by our institutional review board. STAS STAS was defined as tumor cells within air spaces in the lung parenchyma beyond the edge of the main tumor (14, 21). In addition, STAS was also classified into the following three categories: "no STAS" (Figure 1A) for cases without definite STAS, "low STAS" (Figure 1B) for cases with 1 to 4 single cells or clusters of STAS, and "high STAS" (Figure 1C) for cases with five or more single cells or clusters of STAS (13-15). The pathologist ruled out single tumor cells/clusters with broken figure as an artifact because the finding suggests unnaturally separated cells/clusters. All tumors were evaluated at a magnification of x200 using an optical microscope (BX40; Olympus, Tokyo, Japan). Statistical Analyses The associations between STAS and patient characteristics were analysed using Fisher s exact test. The cancer-specific OS was defined as the time from the initial surgery until death from lung cancer, while the recurrence-free survival (RFS) was defined as the time from the initial surgery until recurrence. The Kaplan-Meier method was used to estimate the survival probabilities, and the curves of the two or three groups were statistically compared using the log-rank test. The prognostic factors for the DFS and OS were assessed using a logistic regression model with a backward elimination method. The univariate and multivariate analyses of the relationship between STAS and the preoperative clinicopathological features were performed using logistic 5

7 regression with backward elimination. All of the statistical analyses were conducted using the JMP version 12 software program (SAS Institute, Cary, NC, USA). P values of <0.05 were considered to indicate statistically significant differences. Results Association between STAS and the Clinicopathological Characteristics Table 1 shows the characteristics of the patients included in this translational study. Among 276 patients, 123 (44.6%), 48 (17.4%), and 105 (38.0%) were classified as having no, low, and high STAS, respectively; 153 (55.4%) were positive for STAS. Fisher s exact test demonstrated that positivity for STAS was significantly associated with a higher CEA value (P=0.013), a larger radiological tumor diameter (P=0.008), a higher C/T ratio (P<0.001), a higher SUVmax (P<0.001), a pathologically larger tumor size (P=0.004), the presence of pleural invasion (P=0.027), a histologically invasive type (P<0.001), and recurrence (P<0.001), while STAS was not significantly associated with EGFR mutations or PD-L1 expression (P=0.129 and P=0.872, respectively; Supplemental Tables 1 and 2). Survival Analyses According to the Presence of STAS The median follow-up time after surgery was 5.0 years. Patients with STAS-positive adenocarcinoma exhibited a significantly shorter RFS and cancer-specific OS than those without (P<0.001 and P=0.002, respectively; Figures 2A and 2B). The 5-year RFS in STAS-negative and STAS-positive patients were 89.5% and 71.7%, respectively. The 5-year cancer-specific OS in STAS-negative and STAS-positive patients were 98.3% and 91.3%, respectively. According to a multivariate analysis, positivity for STAS remained an independent prognostic parameter for both the RFS and the cancer-specific OS (hazard ratio [HR] 3.271, 95% confidence interval [CI] , P=0.001 and HR 5.671, 95% CI , P<0.001, respectively; Tables 2 and 3). The differences in the RFS and cancer-specific OS between the no, low, and high STAS were significant as shown in Figures 2C and 2D. In patients who underwent limited resection, patients with STAS-positive adenocarcinoma exhibited a significantly shorter RFS and cancer-specific OS than those without (P<0.001 and P=0.001, respectively; Supplemental figures 1A and 1B). In 6

8 patients who underwent lobectomy or bilobectomy, patients with STAS-positive adenocarcinoma exhibited a significantly shorter RFS (P=0.042; Supplemental Figure 2A) but not cancer-specific OS (P=0.099; Supplemental Figure 2B). Prediction of STAS using Preoperative Parameters A univariate test showed that a higher CEA value (P=0.009), a radiological tumor size 2.0 cm (P=0.007), a higher C/T ratio (P<0.001), and a higher SUVmax (P<0.001) were significantly associated with the presence of STAS (Supplemental Table 3). In multivariate analyses, a higher C/T ratio (P=0.041) and a higher SUVmax (P=0.001) were independent predictors of STAS. Comment In the current study, STAS was observed in 153 (55.4%) among 276 patients and the presence of STAS was significantly associated with the clinicopathologically invasive features, such as a larger tumor size, the presence of pleural invasion and a histologically invasive type. Furthermore, multivariate analyses showed that positivity for STAS remained an independent prognostic parameter for both the RFS and OS. Our findings, as well as those in previous reports (13-15), suggested that small-sized adenocarcinomas with STAS might be a special entity that should be optimally resected and intensely followed up after surgical resection. The prognostic significance of STAS remains controversial in patients who underwent limited resection or lobectomy. First, Kadota et al. reported that STAS was a significant risk factor of recurrence in small-sized adenocarcinomas treated with limited resection but not in those who underwent lobectomy (13). However, according to the reports by Uruga et al. and Warth et al., STAS was significantly associated with a worse RFS and OS in patients with early-stage and stage I-IV adenocarcinoma that had been resected (15, 21). Our results showed that STAS negatively impacted the RFS in both the limited resection and lobectomy groups. With regard to the OS, STAS was significantly associated with a worse OS in the limited resection group, while a nonsignificant trend was found between STAS and the OS in the lobectomy groups. The significance of STAS in small-sized pulmonary adenocarcinomas treated with lobectomy has therefore yet to be confirmed and should be explored further in future prospective studies. STAS-positive small 7

9 adenocarcinomas that undergo limited resection should be carefully followed up after the surgical operation. It appears to be essential to preoperatively identify the presence of STAS since thoracic surgeons would then be able to select the optimal surgical procedures for such patients. The present study showed that among preoperative parameters, a C/T ratio 0.25 and a SUVmax 2.5 were independent predictors of STAS. In addition, Shiono et al. reported that STAS was significantly related to solid nodules on CT (14). These findings suggest that the presence of a solid part and a higher SUVmax might be useful for predicting the presence of STAS. Although Shiono et al. divided their study population into non-solid and pure solid on CT, we investigated the associations between STAS and the C/T ratio in more detail (Supplemental Table 2). Intriguingly, among 153 patients with STAS, 17 (11.1%) patients presented as GGO with a C/T ratio of 0.25, which was defined as radiological noninvasiveness on CT (2). Although limited resection might be sufficient for patients with a radiologically noninvasive nodule, and while several prospective studies have investigated the significance of limited resection for patients with small lung adenocarcinomas harboring some extent of consolidation (22), our results suggest that some radiologically noninvasive nodules should be treated with lobectomy combined with nodal dissection but not limited resection. Future studies should focus on the identification of what kind of radiologically noninvasive lesions harbor STAS. We also plan to analyze CT features, such as convergence, indentation, and cavity, as well as the presence of GGO and solid components. While STAS was observed in 153 (55.4%) of the 276 patients in the present study, in three representative reports that investigated the frequency of STAS in patients with lung adenocarcinoma, the frequency of STAS was 38.0%, 14.8% and 52.4%, respectively (13-15). These differences suggest that the methods by which the pathological slides were prepared differed according to the institution or pathologist. Although we believe that our method of pathological slide preparation is uniformly accepted at a higher level, since this was a retrospective analysis, our protocol (and also the protocol used in the other retrospective studies) was not made for identifying STAS. A multicenter prospective study conducted by Blaauwgeers et al. showed that as many as 93% of loose tissue fragments could be explained by the mechanical forces associated with tissue handling and this was recognized as Spread Through A Knife Space 8

10 (23). Thus, there is a need for a precise pathological slide preparation protocol that minimizes artifacts to the maximum possible extent. Another possible reason for STAS is an artifact during the surgical procedure and surgical handling or stapling techniques might affect STAS formation. In the current study, the frequency of STAS was significantly lower in the limited resection group than in the lobectomy group (P<0.001; Table 1), which was similar to the findings of a previous report (13). In contrast, other reports have found no significant differences in the formation of STAS among surgical procedures (13, 14). This discrepancy among reports might be due to differences in the surgical manipulation among surgeons and institutions, which cannot be standardized. Alternatively, some biological process might induce STAS formation. According to previous reports, mutations in the KRAS and BRAF genes were significantly associated with STAS, and a wild-type EGFR gene was also found to be significantly associated with STAS (14, 21). In the current study, STAS was not significantly associated with the wild-type EGFR gene or the expression of PD-L1, a key molecule in the tumor microenvironment (24). The current study is the first to investigate the relationship between STAS and PD-L1 expression. In esophageal cancer, tumor budding at the tumor invasive front has been reported to be associated with the epithelial to mesenchymal transition (EMT) phenomenon (25), and we plan to analyze the association between STAS and EMT for E-cadherin and vimentin. Furthermore, what still remains a mystery is how and why their pathologists began to recognize this entity now rather than previously. With regard to this point, pathologists might have been aware of STAS to some degree before it was first introduced in 2015 by Kadota et al. (13). However, while those pathologists might have recognized floating cells in alveolar spaces, the clinicopathological significance of such cells had not been investigated; therefore, the significance of this finding was unclear before relevant terminology were established. Regarding the reasons for this lack of recognition, Kadota et al. proposed the following (13): First, this pattern of invasion is unique to the lung compared with other organs because lung anatomy differs due to presence of air spaces, which normally contain air but also provide a path through which tumor cells can spread. Second, STAS is easily dismissed as an artifact where cells within air spaces were regarded as floaters or carry over due to contamination during processing. Because pathologists are not trained to look for these cells in the lung parenchyma beyond the edge of lung 9

11 adenocarcinomas, STAS is mostly overlooked on microscopic review. Furthermore, due to paucity of data, there has been little emphasis on the clinical significance of this finding, and pathologists have not been compelled to look for this routinely. We therefore believe that it will take some time for the STAS phenomenon to be recognized prevalently and hope that our report will contribute to such recognition. Limitations of the current study include its retrospective, single-institutional design and limited sample sizes. Large-scale, prospective studies examining STAS are thus warranted. In conclusion, STAS was associated with clinicopathologically invasive features and it was also predictive of a poor survival. 10

12 References 1. Ginsberg RJ, Rubinstein LV. Randomized trial of lobectomy versus limited resection for t1 n0 non-small cell lung cancer. Lung cancer study group. Ann Thorac Surg 1995;60(3): ; discussion Suzuki K, Koike T, Asakawa T et al. 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 2011;6(4): Rami-Porta R, Bolejack V, Crowley J et al. The iaslc lung cancer staging project: Proposals for the revisions of the t descriptors in the forthcoming eighth edition of the tnm classification for lung cancer. J Thorac Oncol 2015;10(7): Uehara H, Tsutani Y, Okumura S et al. Prognostic role of positron emission tomography and high-resolution computed tomography in clinical stage ia lung adenocarcinoma. Ann Thorac Surg 2013;96(6): Yoshiya T, Mimae T, Tsutani Y et al. Prognostic role of subtype classification in smallsized pathologic n0 invasive lung adenocarcinoma. Ann Thorac Surg 2016;102(5): Travis WD who classification of the pathology and genetics of tumors of the lung. J Thorac Oncol 2015;10(9):S68-S Travis WD, Brambilla E, Noguchi M et al. International association for the study of lung cancer/american thoracic society/european respiratory society international multidisciplinary classification of lung adenocarcinoma. J Thorac Oncol 2011;6(2): Shiono S, Ishii G, Nagai K et al. Histopathologic prognostic factors in resected colorectal lung metastases. Ann Thorac Surg 2005;79(1): ; discussion Shiono S, Ishii G, Nagai K et al. Predictive factors for local recurrence of resected colorectal lung metastases. Ann Thorac Surg 2005;80(3): Onozato ML, Kovach AE, Yeap BY et al. Tumor islands in resected early-stage lung adenocarcinomas are associated with unique clinicopathologic and molecular characteristics and worse prognosis. Am J Surg Pathol 2013;37(2): Thunnissen E, Blaauwgeers HJ, de Cuba EM, Yick CY, Flieder DB. Ex vivo artifacts and histopathologic pitfalls in the lung. Arch Pathol Lab Med 2016;140(3):

13 12. Warth A, Beasley MB, Mino-Kenudson M. Breaking new ground: The evolving concept of spread through air spaces (stas). J Thorac Oncol 2017;12(2): Kadota K, Nitadori J, Sima CS et al. Tumor spread through air spaces is an important pattern of invasion and impacts the frequency and location of recurrences after limited resection for small stage i lung adenocarcinomas. J Thorac Oncol 2015;10(5): Shiono S, Yanagawa N. Spread through air spaces is a predictive factor of recurrence and a prognostic factor in stage i lung adenocarcinoma. Interact Cardiovasc Thorac Surg 2016;23(4): Uruga H, Fujii T, Fujimori S, Kohno T, Kishi K. Semiquantitative assessment of tumor spread through air spaces (stas) in early-stage lung adenocarcinomas. J Thorac Oncol 2017;12(7): Kadota K, Kushida Y, Katsuki N et al. Tumor spread through air spaces is an independent predictor of recurrence-free survival in patients with resected lung squamous cell carcinoma. Am J Surg Pathol Lu S, Tan KS, Kadota K et al. Spread through air spaces (stas) is an independent predictor of recurrence and lung cancer-specific death in squamous cell carcinoma. J Thorac Oncol 2017;12(2): Takenaka T, Takenoyama M, Toyozawa R et al. Concurrent chemoradiotherapy for patients with postoperative recurrence of surgically resected non-small-cell lung cancer. Clin Lung Cancer 2015;16(1): Kohno M, Okamoto T, Suda K et al. Prognostic and therapeutic implications of aromatase expression in lung adenocarcinomas with egfr mutations. Clin Cancer Res 2014;20(13): Takada K, Okamoto T, Shoji F et al. Clinical significance of pd-l1 protein expression in surgically resected primary lung adenocarcinoma. J Thorac Oncol 2016;11(11): Warth A, Muley T, Kossakowski CA et al. Prognostic impact of intra-alveolar tumor spread in pulmonary adenocarcinoma. Am J Surg Pathol 2015;39(6): Aokage K, Saji H, Suzuki K et al. A non-randomized confirmatory trial of segmentectomy for clinical t1n0 lung cancer with dominant ground glass opacity based on thin- 12

14 section computed tomography (jcog1211). Gen Thorac Cardiovasc Surg 2017;65(5): Blaauwgeers H, Flieder D, Warth A et al. A prospective study of loose tissue fragments in non-small cell lung cancer resection specimens: An alternative view to "spread through air spaces". Am J Surg Pathol 2017;41(9): Pardoll DM. The blockade of immune checkpoints in cancer immunotherapy. Nat Rev Cancer 2012;12(4): Tsutsumi S, Saeki H, Nakashima Y et al. Programmed death-ligand 1 expression at tumor invasive front is associated with epithelial-mesenchymal transition and poor prognosis in esophageal squamous cell carcinoma. Cancer Sci 2017;108(6):

15 Figure Legends Figure 1. Histological figures of spread through air spaces (STAS); the pictures show pulmonary adenocarcinoma cases (A) without STAS, (B) with a few floating clusters (arrows) of adenocarcinoma cells (low STAS), and (C) numerous clusters (arrows) of adenocarcinoma cells (high STAS) in the alveolar spaces near the margin of the tumors. Figure 2. Kaplan-Meier curves according to spread through air spaces (STAS). The (A) recurrence-free and (B) cancer-specific overall survivals were significantly worse in patients with STAS than without (P<0.001 and P=0.002, respectively). Among no, low, and high STAS, there were significant differences in the (C) recurrence-free and (D) cancer-specific overall survivals (P<0.001 and P=0.007, respectively). Supplementary Figure 1. Kaplan-Meier curves according to spread through air spaces (STAS) in patients who underwent limited resection. The (A) recurrence-free and (B) cancer-specific overall survivals were significantly worse in patients with STAS than without (P<0.001 and P=0.001, respectively). Supplementary Figure 2. Kaplan-Meier curves according to spread through air spaces (STAS) in patients who underwent lobectomy or bilobectomy. The (A) recurrence-free survival was significantly worse in patients with STAS than without (P=0.042), while there was no significant difference in the (B) cancer-specific overall survival between those with and without STAS (P=0.099). 14

16 Alphabetical list of all abbreviations spread through air spaces: STAS non-small cell lung cancer: NSCLC ground-glass opacity: GGO overall survival: OS computed tomography: CT positron emission tomography: PET carcinoembryonic antigen: CEA tumor-node-metastasis: TNM epidermal growth factor receptor: EGFR programmed cell death-1 ligand: PD-L1 recurrence-free survival: RFS hazard ratio: HR confidence interval: CI epithelial to mesenchymal transition: EMT 15

17 Table 1. Patient characteristics Factors Number of patients (n=276) Age (years) Median 69 Range Sex Female 142 (51.5%) Male 134 (48.5%) Smoking status Never-smoker 139 (50.4%) Smoker 137 (49.6%) CEA* (62.1%) > (37.9%) Radiological tumor (53.6%) diameter (cm) > (46.4%) C/T ratio 0 (pure GGO) 36 (13.0%) (6.9%) (14.5%) (65.6%) SUVmax* < (50.9%) (49.1%) Pathological T status IA 139 (50.4%) IB 67 (24.3%) IIA 70 (25.3%) Pathological stage IA 206 (74.6%) IB 70 (25.4%) pl* Absent 241 (87.6%) Present 34 (12.4%) ly Absent 257 (93.1%) Present 19 (6.9%) v Absent 229 (83.0%) Present 47 (17.0%) Histological subtype AAH/AIS/MIA 35 (12.7%) Lepidic predominant 20 (7.2%) Papillary predominant 198 (71.7%) Acinar predominant 3 (1.1%) Micropapillary predominant 1 (0.4%) Solid predominant 14 (5.1%) Variants 5 (1.8%) Surgical procedure Wedge resection 54 (19.6%) Segmentectomy 35 (12.7%) Lobectomy 184 (66.6%) Bilobectomy Pneumonectomy 1 (1.1%) 0 (0.0%) 16

18 Adjuvant chemotherapy None 239 (86.6%) UFT 37 (13.4%) EGFR* Wildtype 90 (52.0%) Mutant 83 (48.0%) PD-L1 expression Negative 229 (83.0%) Positive 47 (17.0%) Type of recurrence None 235 (85.1%) Local 21 (7.6%) Distant 20 (7.3%) *: cases in which data were available. CEA: carcinoembryonic antigen, C/T: consolidation/tumor, SUVmax: maximum standardized uptake value, pl: pleural invasion, ly: lymphatic invasion, v: vascular invasion, EGFR: epidermal growth factor receptor, GGO: ground-glass opacity, AAH: atypical adenomatous hyperplasia, AIS: adenocarcinoma in situ, MIA: minimally invasive adenocarcinoma, PD-L1: programmed death-ligand 1, UFT: tegafur-uracil. 17

19 Table 2. Univariate and multivariate analyses of the relationship between the disease-free survival and the clinicopathological factors Factors Univariate analysis Multivariate analysis HR 95% CI P value HR 95% CI P value Age (years) < Sex Female Male Smoking Never-smoker history Smoker CEA* > < <0.001 Radiological tumor diameter (cm) > SUVmax* < <0.001 Pathological IA stage IB <0.001 pl* Negative Positive < <0.001 ly Negative Positive <

20 v Negative Positive Histological AAH/AIS/MIA subtype Others** Surgical procedure Sublobar resection Lobectomy or bilobectomy Adjuvant None chemotherapy UFT EGFR* Wildtype Mutant PD-L1 Negative Positive STAS Negative Positive < *: cases in which data were available. **: lepidic, papillary, acinar, micropapillary and solid predominant and variants. 19

21 HR: hazard ratio, CI: confidence interval, CEA: carcinoembryonic antigen, SUVmax: maximum standardized uptake value, pl: pleural invasion, ly: lymphatic invasion, v: vascular invasion, AAH: atypical adenomatous hyperplasia, AIS: adenocarcinoma in situ, MIA: minimally invasive adenocarcinoma, UFT: tegafururacil, EGFR: epidermal growth factor receptor, PD-L1: programmed death-ligand 1, STAS: spread through air spaces. 20

22 Table 3. Univariate and multivariate analyses of the relationship between the cancer-specific overall survival and the clinicopathological factors Factors Univariate analysis Multivariate analysis HR 95% CI P value HR 95% CI P value Age (years) < <0.001 Sex Female Male Smoking Never-smoker history Smoker CEA* > < <0.001 Radiological tumor diameter (cm) > SUVmax* < Pathological IA stage IB < <0.001 pl Negative Positive ly Negative Positive

23 v Negative Positive <0.001 Histological AAH/AIS/MIA subtype Others** Surgical procedure Sublobar resection Lobectomy or bilobectomy Adjuvant None chemotherapy UFT EGFR* Wildtype Mutant PD-L1 Negative Positive STAS Negative Positive <0.001 *: cases in which data were available. **: lepidic, papillary, acinar, micropapillary and solid predominant and variants. 22

24 HR: hazard ratio, CI: confidence interval, CEA: carcinoembryonic antigen, SUVmax: maximum standardized uptake value, pl: pleural invasion, ly: lymphatic invasion, v: vascular invasion, AAH: atypical adenomatous hyperplasia, AIS: adenocarcinoma in situ, MIA: minimally invasive adenocarcinoma, UFT: tegafururacil, EGFR: epidermal growth factor receptor, PD-L1: programmed death-ligand 1, STAS: spread through air spaces. 23

25

26