Radiologic findings to predict low-grade malignant tumour among clinical T1bN0 lung adenocarcinomas: lessons from histological subtypes

Japanese Journal of Clinical Oncology, 2015, 45(8) 767 773 doi: 10.1093/jjco/hyv078 Advance Access Publication Date: 7 June 2015 Original Article Original Article Radiologic findings to predict low-grade malignant tumour among clinical T1bN0 lung adenocarcinomas: lessons from histological subtypes Takahiro Mimae 1, Yoshihiro Miyata 1, Takeshi Mimura 1, Hiroyuki Ito 2, Haruhiko Nakayama 2, Sakae Okumura 3, Masahiro Yoshimura 4, and Morihito Okada 1, * 1 Department of Surgical Oncology, Hiroshima University, Hiroshima, 2 Department of Thoracic Surgery, Kanagawa Cancer Centre, Yokohama, 3 Department of Thoracic Surgery, Cancer Institute Hospital, Tokyo, and 4 Department of Thoracic Surgery, Hyogo Cancer Centre, Akashi, Japan *For reprints and all correspondence: Morihito Okada, Department of Surgical Oncology, Hiroshima University, 1-2-3 Kasumi, Minami-Ku, Hiroshima 734-8551, Japan. E-mail: morihito1217@hiroshima-u.ac.jp Received 23 January 2015; Accepted 28 April 2015 Abstract Objective: Some clinical T1bN0 (ct1bn0) lung adenocarcinomas (2 3 cm) are thought to have lessaggressive and less-malignant behaviour although most ct1an0 tumours ( 2 cm) are indolent. The present study aimed to identify pre-operative radiographic findings that can predict ct1bn0 lung adenocarcinoma with low-malignant aggressiveness in consideration of histological subtypes. Methods: The clinicopathological features and prognoses of 224 consecutive patients (histological subtype set, n = 122; prognosis set, n = 224) with ct1bn0 lung adenocarcinoma were retrospectively examined. Adenocarcinoma in situ, minimally invasive adenocarcinoma, lepidic, node-negative papillary and node-negative acinar predominant invasive adenocarcinomas were defined as lowgrade malignant, whereas solid, micropapillary, node-positive acinar and node-positive papillary predominant invasive adenocarcinoma were defined as high-grade malignant. Results: Receiver operating characteristics analysis revealed that the criteria of solid tumour size 1.8 cm on high-resolution computed tomography and the maximum standardized uptake value 3.2 on positron emission tomography/computed tomography could predict low-grade malignant tumour in the histological subtype set. Among 95 (42.4%) of 224 patients who met the criteria for the prognosis set, 94 (98.9%) had no lymph node metastasis and 93 (97.9%) had no recurrence (median follow-up, 43.6 months). The 3 year recurrence-free survival rates were 94.9 and 79.0% in patients whose pre-operative findings met and did not meet the criteria, respectively. Conclusions: Pre-operative radiographic findings of solid tumour size and the maximum standardized uptake value could identify low-grade malignant tumour among ct1bn0 lung adenocarcinomas, which account for about half of all ct1bn0 tumours. Patients with pre-operative lung tumour findings that fulfill the criteria could be candidates for sublobar resection. Key words: solid tumour size, SUV max, sublobar resection, lung adenocarcinoma, subtypes The Author 2015. Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oup.com 767

768 Low-grade lung tumour radiologic findings Introduction Lung adenocarcinoma has been subtyped according to a novel International Association for the Study of Lung Cancer/American Thoracic Society/European Respiratory Society (IASLC/ATS/ERS) classification that was proposed in 2011 (1). Lung adenocarcinoma can be stratified according to malignant aggressiveness as indolent (adenocarcinoma in situ, minimally invasive adenocarcinoma and lepidic predominant invasive adenocarcinoma), intermediate (papillary and acinar predominant invasive adenocarcinoma) and highly aggressive (solid and micropapillary predominant invasive adenocarcinoma) based on a study of prognosis according to histological subtypes (2 6). Papillary and acinar predominant invasive carcinomas comprise the majority of lung adenocarcinomas but the malignant behaviour of these subtypes is highly heterogeneous with both indolent and aggressive phenotypes, whereas that of the other subtypes is rarely heterogeneous (2). The prognosis should be better for lung cancer patients without, than with lymph node metastasis. Thus lung adenocarcinoma might be separated into a low-grade malignant subtype comprising adenocarcinoma in situ, minimally invasive adenocarcinoma, lepidicpredominant invasive adenocarcinoma, lymph node-negative papillary and node-negative acinar predominant invasive adenocarcinoma, and a high-grade malignant subtype comprising solid, micropapillary, lymph node-positive papillary and node-positive acinar predominant invasive adenocarcinoma. To which extent lung resection is adequate as a curative treatment for clinical T1bN0 (ct1bn0) lung adenocarcinoma is controversial because this population has heterogeneous malignant aggressiveness, even though ct1an0 tumours are regarded as a candidate for sublobar resection because they have an indolent phenotype (7,8). That sublobar resection and lobectomy are adequate for indolent and aggressive cancer, respectively, is a reasonable supposition. Thus, preoperative criteria that can predict the indolent phenotype of ct1bn0 lung adenocarcinoma are urgently needed. The criteria for predicting low-grade malignancy have the potential to select candidates for sublobar resection, thus, the criteria should be defined according to both solid tumour size excluding a groundglass opacity (GGO) component on high-resolution computed tomography (HRCT) and the maximum standard uptake value (SUV max )on 2-[18F]fluoro-2-deoxy-D-glucose-positron emission tomography/ computed tomography (FDG-PET/CT) because these parameters indicate the malignant potential of lung adenocarcinoma more accurately than whole tumour size (9 14). Here, we aimed to define which pre-operative radiographic findings including solid tumour size and SUV max could predict low-grade malignant tumour in a histological subtype set that completely included histological subtype information in two institutions. We also investigated a prognosis set that included additional data obtained from two other institutions using the radiographic findings defined above to validate whether the criteria can predict favourable candidates for sublobar resection among patients with ct1bn0 lung adenocarcinoma. The present results will provide useful information and facilitate to determine the strategy of treatment for ct1bn0 lung adenocarcinoma. Patients and methods Patient population The histological subtype set comprised 122 consecutive patients with complete pathological data who underwent surgically curative R0 resection of ct1bn0 lung adenocarcinoma at Hiroshima University Hospital and Kanagawa Cancer Centre between April 2006 and December 2010. The prognosis set comprised above 122 patients and additional 102 consecutive patients (total 224) with complete clinicopathological data except for histological subtype who underwent surgically curative R0 resection of ct1bn0 lung adenocarcinoma at the Cancer Institute Hospital (Tokyo, Japan) and Hyogo Cancer Centre between April 2007 and December 2010. The Institutional Review Boards of the participating institutions approved this retrospective review of a prospective database and waived the requirement for informed consent from individual patients. All patients were staged according to the TNM Classification of Malignant Tumours, 7th edition (15). Endobronchial ultrasonography or mediastinoscopy was not routinely performed. Lymph node metastasis is determined as negative when swollen mediastinal or hilar lymph nodes were not evident on HRCT, and FDG did not accumulate in these lymph nodes according to FDG-PET. Patients who underwent wedge resection without lymph node assessment were excluded from analysis. Segmentectomy with hilar and mediastinal lymph node dissection was generally allowed for GGO-predominant (consolidation-to-whole ratio: 0.5) GGO-mixed tumours, which a previous study regarded as node negative and non-invasive (16). If any lymph node involvement was detected on intraoperative frozen sections, the procedure was converted to standard lobectomy. All other patients underwent standard lobectomy. The inclusion criteria included pre-operative staging determined by HRCT and FDG-PET/CT and curative surgery without induction therapy. Patients with incompletely resected tumours (R1 or R2) and those with multiple tumours or previous lung surgery were excluded. Pathological studies The surgically resected specimens were routinely fixed in 10% buffered formalin, cut serially into 5 7 mm thick slices, and macroscopically assessed. Hematoxylin eosin (H E) stained tissue was histologically diagnosed. Immunohistochemical staining proceeded if a diagnosis could not be established by H E staining alone. Visceralpleural invasion was taken as either absent or present, and at least one tumour cell cluster visible in a lymphatic or blood vessel was considered to indicate lymphatic or blood vessel invasion, respectively. In addition, lymphatic and blood vessel invasion was immunohistochemically assessed by D2-40 and Elastica-Van Gieson (EVG) staining and pleural invasion was evaluated by staining elastic tissue fibres with EVG as necessary. According to the IASLC/ATS/ERS classification, each tumour was reviewed by comprehensive histologic subtyping, and the ratio (%) of each histological component was recorded in 5% increments (1). The predominant subtype with the highest occupancy in each tumour was determined as lepidic, including adenocarcinoma in situ and minimally invasive adenocarcinoma; acinar; papillary; micropapillary or solid. Pathologists from each institution histologically assessed all specimens (1). HRCT imaging Chest images were acquired using 16-row multidetector CT. Highresolution images of the tumours were acquired using the following parameters: 120 kvp, 200 ma, 1 2 mm section thickness, 512 512 pixel resolution, 0.5 1.0 s scanning time, a high spatial reconstruction algorithm with a 20 cm field of view and mediastinal [40 Hounsfield units (HU); width, 400 HU] and lung ( 600 HU; width, 1600 HU) window settings. We defined solid tumour size as the maximum dimension of the solid component of the lung windows, excluding GGO (12), which we defined as a hazy increase in lung attenuation

Jpn J Clin Oncol, 2015, Vol. 45, No. 8 769 that did not obscure underlying vascular markings and the solid component was defined as having residual high density. Radiologists from each institution reviewed all CT images and determined tumour sizes. FDG-PET/CT imaging The patients fasted for at least 4 h before receiving an intravenous injection of 74 370 MBq of FDG and then relaxed for at least 1 h before FDG-PET/CT image acquisition using a Biograph Sensation 16 (Siemens Healthcare, Erlangen, Germany), Aquiduo (Toshiba Medical Systems Corporation, Tochigi, Japan) or Discovery ST (GE Healthcare, Little Chalfont, UK) integrated three-dimensional PET/CT scanners. Low-dose non-enhanced CT images of 2 4 mm section thickness were taken from the head to the pelvis of each patient. Variations in standardized uptake values (SUVs) among the institutions were minimized using an anthropomorphic body phantom (NEMA NU2-2001; Data Spectrum Corp, Hillsborough, NC, USA) (10,17). The original SUV max values were determined by radiologists from each institution for the purposes of this study. On FDG-PET/CT images, all lymph nodes in the thorax with FDG uptake no greater than the normal background activity of the mediastinal blood pool (SUV max <1.5) regardless of size, were considered as cn0. A lymph node where the SUV max was 1.5 was considered suspected malignancy. However, even lymph nodes with high FDG uptake were also considered benign when they had higher attenuation than mediastinal structures (great vessels) or benign central, nodular, diffuse or popcorn-like calcification (18). Follow-up evaluation All patients were followed up by a physical examination and chest radiography every 3 months, as well as by chest and abdominal CT examinations every 6 months for the first 2 years from the day after Table 1. Clinicopathological findings among patients with clinical T1bN0M0 tumours in histological subtype set stratified according to pathological tumour grade Variables All (n = 122) Low grade (n = 99) High grade (n = 23) P Age Median 69 69 64 0.20 Range 37 89 37 79 38 83 Sex Female 69 (57%) 57 (58%) 12 (52%) 0.65 Male 53 (43%) 42 (42%) 11 (48%) CEA (ng/ml) Median 2.7 2.5 3.45 0.74 Range 1.0 18.0 1.0 18.0 1.1 8.4 Whole tumour size (cm) a Median 2.5 2.5 2.5 0.90 Range 2.1 3.0 2.1 3.0 2.1 3.0 Solid tumour size (cm) a Median 1.8 1.4 2.4 <0.001 Range 0 3.0 0 3.0 0.96 3.0 b SUV max Median 1.8 1.6 6.0 <0.001 Range 0 12.4 0 12.4 1.5 10.8 Histological subtypes <0.001 Adenocarcinoma in situ 10 (8%) 10 (10% 0 (0%) Minimally invasive adenocarcinoma 2 (2%) 2 (2%) 0 (0%) Lepidic predominant 42 (34%) 42 (42%) 0 (0%) Papillary predominant 34 (28%) 30 (30%) c 4 (17%) d Acinar predominant 17 (14%) 15 (15%) c 2 (9%) d Solid predominant 13 (11%) 0 (0%) 13 (57%) Micropapillary predominant 4 (3%) 0 (0%) 4 (17%) Pathological tumour size (cm) Median 2.5 2.4 2.5 0.53 Range 1.2 6.0 1.2 6.0 1.5 4.3 Lymphatic invasion Positive 20 (16%) 10 (10%) 10 (43%) <0.001 Blood vessel invasion Positive 27 (22%) 12 (12%) 15 (65%) <0.001 Pleural invasion Positive 19 (16%) 9 (9%) 10 (43%) <0.001 Lymph node metastasis Positive 9 (7%) 0 (0%) 9 (39%) <0.001 a Tumour size determined by high-resolution computed tomography. b Maximum standardized uptake value. c Node negative. d Node positive.

770 Low-grade lung tumour radiologic findings lung resection. Thereafter, patients were assessed by a physical examination and chest radiography every 6 months and annual chest CT. Statistical analyses Continuous and categorical variables were analysed using the Mann Whitney U-test and the χ2 or Fisher s exact tests. Receiver operating characteristics (ROC) curves of solid tumour size and SUV max for predicting low-grade malignant group were generated to determine a cutoff value that maximizes the sum of sensitivity and specificity. Recurrence-free survival (RFS) curves were calculated using the Kaplan Meier method and the two groups were compared using univariate log-rank analyses. All data were statistically analysed using EZR (Saitama Medical Centre, Jichi Medical University), which is a graphical user interface for R (The R Foundation for Statistical Computing, version 2.13.0) (19). More precisely, it is a modified version of R commander version 1.6-3 that was designed to add statistical functions frequently used in biostatistics. P 0.05 was deemed to indicate statistical significance. Results Clinicopathological findings from patients with ct1b lung adenocarcinoma in the histological subtype set Table 1 shows that the median age of the histological subtype set was 69 years and that it comprised more female (n = 69; 57%) than male (n = 53; 43%) patients. The median level of pre-operative carcinoembryonic antigen (CEA) was essentially normal (2.7 ng/ml). The median sizes of whole and solid tumours were 2.5 and 1.8 cm, respectively. The SUV max was not very high (median: 1.8). Histological subtypes consisted of 54 lepidic adenocarcinomas (10 adenocarcinoma in situ, 2 minimally invasive adenocarcinoma and 42 lepidicpredominant invasive adenocarcinoma), 34 papillary, 17 acinar, 13 solid and 4 micropapillary predominant invasive adenocarcinoma. The median pathological tumour size (2.5 cm) was similar to the whole tumour size on pre-operative HRCT. Only one tumour whose pathological size was 6 cm was recognized. Lymphatic, blood vessel and pleural invasion and lymph node metastasis were detected in 20 (16%), 27 (22%), 19 (16%) and 9 (7%) patients, respectively. The histological subtype set included 18 sublobar resections (all segmentectomies) and 104 lobectomies. Clinicopathological findings from patients in the histological subtype set with ct1b lung adenocarcinoma classified according to low- and high-grade malignancy The low- and high-grade malignant groups included 99 and 23 patients, respectively. Age (69 vs. 64 years), sex (female, n = 57, 58% vs. n = 12, 52%), CEA (2.5 vs. 3.45 ng/ml) and whole tumour size (2.5 vs. 2.5 cm) did not significantly differ between the groups whereas solid tumour size and SUV max were significantly higher in highthan the low-grade group (2.4 vs. 1.4 cm and 6.0 vs. 1.6, both P < 0.001). The positive rates of lymphatic, blood vessel and pleural invasion and lymph node metastasis were significantly higher in the high- than in the low-grade group (Table 1). Pre-operative radiographic findings for predicting low-grade malignancy in the histological subtype set using ROC analysis The optimal cut-offs for predicting low-grade malignant tumour among patients with ct1bn0 lung adenocarcinoma determined by ROC analysis were a solid tumour size of 1.8 cm on HRCT and an SUV max of 3.2 on PET/CT (Fig. 1). The clinicopathological findings were then analysed using solid tumour size 1.8 cm and/or SUV max 3.2. Clinicopathological findings and RFS according to criteria for predicting low-grade malignancy in the prognosis set We analysed the prognosis set using the above criteria. Lymph node metastasis was evident in one (0.9%) and nine (5.8%) patients whose tumours met the criteria of solid tumour size 1.8 cm and/or SUV max 3.2, respectively (Table 2). The prognosis set included 27 sublobar resections (all segmentectomies) and 197 lobectomies. The median follow-up was 43.6 months. The 3-year-RFS rate for all included patients was 85.6% [95% confidence interval (CI), Figure 1. Receiver operating characteristics (ROC) curves of solid tumour size and the maximum standardized uptake value (SUV max ) for predicting low-grade malignancy among patients with clinical T1bN0 (ct1bn0) lung adenocarcinoma. (A) Optimal cut-off of solid tumour size for ct1bn0 lung adenocarcinoma, 1.8 cm (n = 122; area under the curve (AUC), 0.82; 95% confidence interval (CI), 0.73 0.90; P < 0.001). (B) Optimal SUV max cut-off for ct1bn0 lung adenocarcinoma, 3.2 (n = 122; AUC, 0.86; 95% CI, 0.79 0.93; P < 0.001).

Jpn J Clin Oncol, 2015, Vol. 45, No. 8 771 Table 2. Clinicopathological findings among patients in prognosis set with clinical T1bN0M0 tumours stratified according to low-grade malignant tumour criteria of solid tumour size 1.8 cm and SUV max 3.2 Variable Fulfilled (n = 95) Unfulfilled (n = 129) Age Median 67 68 0.57 Range 37 89 33 86 Sex Female 57 (60%) 68 (53%) 0.34 Male 38 (40%) 61 (47%) CEA (ng/ml) Median 2.2 3.0 0.015 Range 0.6 24.3 0.6 24.9 Whole tumour size (cm) a Median 2.4 2.6 0.0012 Range 2.1 3 2.1 3 Solid tumour size (cm) a Median 0.75 2.4 <0.001 Range 0 1.78 0.88 3 b SUV max Median 1.3 3.6 <0.001 Range 0 2.9 0.7 12.4 Lepidic component ratio Median 0.6 0.2 <0.001 Range 0.1 1 0 1 Lymphatic invasion Positive 2 (2%) 40 (31%) <0.001 Blood vessel invasion Positive 2 (2%) 51 (40%) <0.001 Pleural invasion Positive 1 (1%) 26 (20%) <0.001 Lymph node metastasis Positive 1 (1%) 21 (16%) <0.001 Recurrence Positive 2 (2%) 25 (19%) <0.001 a Determined by high-resolution computed tomography. b Maximum standardized uptake value. P 79.9 89.8%] (Fig. 2A). These rates were 94.9% (95% CI, 86.8 98.1%) and 79.0% (95% CI, 70.4 85.3%) in patients whose preoperative findings fulfilled and did not fulfil the criteria of solid tumour size 1.8 cm and SUV max 3.2, respectively (n =95 and 129, P < 0.001) (Fig. 2B), and 94.1% (95% CI, 88.5 97.0%) and 67.8% (95% CI, 54.9 77.7%) in patients whose pre-operative findings fulfilled and did not fulfil the criteria of solid tumour size 1.8 cm or SUV max 3.2, respectively (n = 154 and 70, P <0.001). Two and seven of the patients who met the former and latter criteria, respectively, had recurrence. One of the two patients with tumours, who met the criterion of solid tumour size 1.8 cm and SUV max 3.2, underwent lobectomy and had distant, including cerebral metastasis. The other who was treated by segmentectomy had peritoneal dissemination. Neither of them had lymph node metastasis. Discussion Here, we identified criteria that can be used to predict low-grade malignant lung adenocarcinoma comprising lepidic and node-negative papillary or acinar predominant invasive adenocarcinoma. Solid tumour size 1.8 cm on HRCT and SUV max 3.2 on PET/CT are proposed as criteria to discriminate low-grade malignant tumour in patients with ct1bn0 lung adenocarcinoma. The criteria of solid tumour size 1.8 cm and SUV max 3.2 and that of solid tumour size 1.8 cm or SUV max 3.2 were candidates for indicating low-grade malignant tumour determined from ROC curves. The present study aimed to determine which pre-operative radiographic findings that can predict favourable outcomes and less-aggressive tumours among patients with ct1bn0 lung adenocarcinoma who could be candidates for sublobar resection. From this viewpoint, solid tumour size 1.8 cm and SUV max 3.2 are more effective. Fewer lymph node metastases and recurrences were detected in patients who met the criteria of solid tumour size 1.8 cm and SUV max 3.2 than the other criteria. The RFS was slightly better among patients who met the criteria of solid tumour size 1.8 cm and SUV max 3.2 compared with 1.8 cm or SUV max 3.2, Figure 2. Kaplan Meier recurrence free survival (RFS) curves for patients with clinical T1bN0 lung adenocarcinoma that fulfilled and did not fulfil radiographic criteria. RFS curves in prognosis set for all patients (A) and for patients who met or did not meet the criteria of solid tumour size 1.8 cm and SUV max < 3.2 (B). RFS curves of patients classified according to each criterion significantly differ (all P < 0.001; log-rank test). Solid and dashed lines indicate patients who fulfilled and did not fulfill each criterion, respectively.

772 Low-grade lung tumour radiologic findings respectively, in the prognosis set. In addition, all sites of recurrence in patients with tumours that met the criteria of solid tumour size 1.8 cm and SUV max 3.2 were distant. Therefore, the surgical procedure did not cause the recurrence. The criteria of solid tumour size >1.8 cm and SUV max > 3.2 are useful for predicting a poor prognosis, although this was not the purpose of the study. Patients with tumours that met these criteria were classified as having high-grade malignant ct1bn0 lung adenocarcinoma. In addition, the rates of pathologically malignant factors including lymphatic and blood vessel invasion, lymph node metastasis and recurrence were much higher in such patients than in those with tumours that met the criteria of solid tumour size 1.8 cm and SUV max 3.2. Post-operative adjuvant chemotherapy comprising oral tegafur uracil is recommended in Japan for patients with pathological T1bN0 lung adenocarcinoma because oral tegafur uracil for 2 years has proven beneficial in that population (20 22). Although the present study did not address this matter, adjuvant therapy could perhaps be omitted for treating tumours that meet the criteria for low-grade malignancy, but recommended post-operatively who those that meet the criteria for high-grade malignancy. Whether or not ct1bn0 non-small cell lung cancer could be taken as early-stage disease and thus a candidate for sublobar resection is controversial (23 25). Presently, JCOG1211 is an on-going clinical Phase II trial of segmentectomy with lymph node dissection for patients with ct1n0 lung cancer that comprises <50% GGO in Japan. The prognosis does not seem to differ after lobectomy or segmentectomy for patients with ct1bn0 lung adenocarcinoma that meets the N0 criteria including solid tumour size 0.8 cm and SUV max 1.5 (26). Based on these findings, malignant aggression indicated as solid tumour size and SUV max can predict prognosis more accurately than the extent of lung resection, lobectomy and sublobar resection in patients with ct1bn0 lung adenocarcinoma. Thus, the selection of the extent of lung resection due to the malignant grade of tumours is critical. We speculate that sublobar resection and lobectomy should be selected to treat low- and high-grade malignant lung adenocarcinoma, respectively, based on histological subtypes. This speculation requires validation in a prospective clinical trial. Conclusions In summary, we propose that the pre-operative radiographic criteria of solid tumour size 1.8 cm and SUV max 3.2 indicating low-grade malignant tumour can histologically predict lepidic adenocarcinoma, node-negative papillary and node-negative acinar predominant invasive adenocarcinoma among patients with ct1bn0 lung adenocarcinoma. Patients with lung lesions whose pre-operative findings fulfill the criteria for low-grade malignancy could be cured by sublobar resection. Funding This study was supported by the Japan Society for the Promotion of Science Kakenhi (26861122 to T.M. and 24390329 to M.O.). Conflict of interest statement None declared. References 1. 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;6:244 85. 2. Okada M. Subtyping lung adenocarcinoma according to the novel 2011 IASLC/ATS/ERS classification: correlation with patient prognosis. Thorac Surg Clin 2013;23:179 86. 3. Sakurai H, Asamura H, Miyaoka E, et al. Differences in the prognosis of resected lung adenocarcinoma according to the histological subtype: a retrospective analysis of Japanese lung cancer registry data. Eur J Cardiothorac Surg 2014;45:100 7. 4. Yanagawa N, Shiono S, Abiko M, Ogata SY, Sato T, Tamura G. New IASLC/ATS/ERS classification and invasive tumor size are predictive of disease recurrence in stage I lung adenocarcinoma. J Thorac Oncol 2013;8:612 8. 5. Yoshizawa A, Motoi N, Riely GJ, et al. Impact of proposed IASLC/ATS/ ERS classification of lung adenocarcinoma: prognostic subgroups and implications for further revision of staging based on analysis of 514 stage I cases. Mod Pathol 2011;24:653 64. 6. Warth A, Muley T, Meister M, et al. The novel histologic International Association for the Study of Lung Cancer/American Thoracic Society/ European Respiratory Society classification system of lung adenocarcinoma is a stage-independent predictor of survival. JClinOncol2012;30:1438 46. 7. Okada M, Koike T, Higashiyama M, Yamato Y, Kodama K, Tsubota N. Radical sublobar resection for small-sized non-small cell lung cancer: a multicenter study. J Thorac Cardiovasc Surg 2006;132:769 75. 8. Tsubota N, Ayabe K, Doi O, et al. Ongoing prospective study of segmentectomy for small lung tumors. Study Group of Extended Segmentectomy for Small Lung Tumor Ann Thoracic Surg 1998;66:1787 90. 9. Mimae T, Miyata Y, Tsutani Y, et al. What are the radiologic findings predictive of indolent lung adenocarcinoma? Jpn J Clin Oncol 2015;45: 367 72. 10. Nakayama H, Okumura S, Daisaki H, et al. Value of integrated positron emission tomography revised using a phantom study to evaluate malignancy grade of lung adenocarcinoma: a multicenter study. Cancer 2010;116:3170 7. 11. Okada M, Nakayama H, Okumura S, et al. Multicenter analysis of highresolution computed tomography and positron emission tomography/computed tomography findings to choose therapeutic strategies for clinical stage IA lung adenocarcinoma. J Thorac Cardiovasc Surg 2011;141:1384 91. 12. Tsutani Y, Miyata Y, Nakayama H, et al. Prediction of pathologic nodenegative clinical stage IA lung adenocarcinoma for optimal candidates undergoing sublobar resection. J Thorac Cardiovasc Surg 2012;144:1365 71. 13. Tsutani Y, Miyata Y, Nakayama H, et al. Prognostic significance of using solid versus whole tumor size on high-resolution computed tomography for predicting pathologic malignant grade of tumors in clinical stage IA lung adenocarcinoma: a multicenter study. JThoracCardiovascSurg2012;143:607 12. 14. Tsutani Y, Miyata Y, Yamanaka T, et al. Solid tumors versus mixed tumors with a ground-glass opacity component in patients with clinical stage IA lung adenocarcinoma: prognostic comparison using high-resolution computed tomography findings. J Thorac Cardiovasc Surg 2013;146:17 23. 15. Goldstraw P, Crowley J, Chansky K, et al. The IASLC Lung Cancer Staging Project: proposals for the revision of the TNM stage groupings in the forthcoming (seventh) edition of the TNM Classification of malignant tumors. J Thorac Oncol 2007;2:706 14. 16. Tsutani Y, Miyata Y, Nakayama H, et al. Appropriate sublobar resection choice for ground glass opacity-dominant clinical stage IA lung adenocarcinoma: wedge resection or segmentectomy. Chest 2014;145:66 71. 17. Delbeke D, Coleman RE, Guiberteau MJ, et al. Procedure guideline for tumor imaging with 18F-FDG PET/CT 1.0. J Nucl Med 2006;47:885 95. 18. Li L, Ren S, Zhang Y, et al. Risk factors for predicting the occult nodal metastasis in T1 2N0M0 NSCLC patients staged by PET/CT: potential value in the clinic. Lung Cancer 2013;81:213 7. 19. Mimae T, Tsutani Y, Miyata Y, et al. Role of lymphatic invasion in the prognosis of patients with clinical node-negative and pathologic

Jpn J Clin Oncol, 2015, Vol. 45, No. 8 773 node-positive lung adenocarcinoma. J Thorac Cardiovasc Surg 2014;147: 1820 6. 20. Hamada C, Tanaka F, Ohta M, et al. Meta-analysis of postoperative adjuvant chemotherapy with tegafur-uracil in non-small-cell lung cancer. J Clin Oncol 2005;23:4999 5006. 21. Kato H, Ichinose Y, Ohta M, et al. A randomized trial of adjuvant chemotherapy with uracil tegafur for adenocarcinoma of the lung. N Engl J Med 2004;350:1713 21. 22. Wada H, Hitomi S, Teramatsu T. Adjuvant chemotherapy after complete resection in non-small-cell lung cancer. West Japan Study Group for Lung Cancer Surgery. J Clin Oncol 1996;14:1048 54. 23. Iwata H. Minimally invasive pulmonary surgery for lung cancer, up to date. Gen Thorac Cardiovasc Surg 2013;61:449 54. 24. Ohtaki Y, Shimizu K. Anatomical thoracoscopic segmentectomy for lung cancer. Gen Thorac Cardiovasc Surg 2014;62:586 93. 25. Sawabata N. Locoregional recurrence after pulmonary sublobar resection of non-small cell lung cancer: can it be reduced by considering cancer cells at the surgical margin? Gen Thorac Cardiovasc Surg 2013;61: 9 16. 26. Tsutani Y, Miyata Y, Nakayama H, et al. Sublobar resection for lung adenocarcinoma meeting node-negative criteria on preoperative imaging. Ann Thorac Surg 2014;97:1701 7.