Ever since Cahan 1 first introduced lymph node dissection

Original Article Mediastinal Nodal Involvement in Patients with Clinical Stage I Non Small-Cell Lung Cancer Possibility of Rational Lymph Node Dissection Tomohiro Haruki, MD,* Keiju Aokage, MD,* Tomohiro Miyoshi, MD,* Tomoyuki Hishida, MD,* Genichiro Ishii, MD, Junji Yoshida, MD,* Masahiro Tsuboi, MD,* Hiroshige Nakamura, MD, and Kanji Nagai, MD* Background: The aim of this study is to elucidate the optimal candidate of selective lymph node dissection (LND) that reduces the extent of mediastinal LND according to clinical information including radiological evaluation in primary non small-cell lung cancer (NSCLC) patients. Methods: Eight hundred and seventy-six patients with clinical(c)-stage I NSCLC (adenocarcinoma and squamous cell carcinoma), who underwent complete surgical resection between January 2003 and December 2009 were included in this study. We elucidated the lymph node metastatic incidence and distribution according to the primary tumor lobe location and extracted the associated clinicopathological factors, especially thin-section computed tomographic findings, with mediastinal lymph node involvement. Results: The total incidence of mediastinal lymph node metastasis was 9.1%. There were no cases with hilar and mediastinal lymph node metastasis in ground glass opacity-predominant tumors. There was no significant association of clinical factors with subcarinal lymph node metastasis in right upper-lobe and left upper-division adenocarcinoma. An elevated preoperative serum carcinoembryonic antigen level (p < 0.001) showed significant associations with upper mediastinal lymph node metastasis in the patients with bilateral lower-lobe primary lung adenocarcinoma. Conclusions: It would be acceptable to perform selective LND in patients with c-stage I NSCLC with ground glass opacitypredominant tumor. Elevated serum carcinoembryonic antigen was associated with upper mediastinal lymph node involvement in lower-lobe primary lung adenocarcinoma with radiologically solid-predominant tumor. We should be careful when applying selective LND to patients with solid-predominant tumor, especially located in the lower lobe. *Division of Thoracic Surgery, Division of Pathology, Research Center for Innovative Oncology, National Cancer Center Hospital East, Chiba, Japan; and Faculty of Medicine, Division of General Thoracic Surgery, Department of Surgery, Tottori University, Tottori, Japan. Disclosure: The authors declare no conflict of interest. Address for correspondence: Keiju Aokage, MD, Division of Thoracic Surgery, National Cancer Center Hospital East, 6-5-1, Kashiwanoha, Kashiwa, Chiba 277-8577, Japan. E-mail: kaokage@east.ncc.go.jp DOI: 10.1097/JTO.0000000000000546 ISSN: 1556-0864/15/1006-0930 Key Words: Non small-cell lung cancer, Mediastinal lymph node metastasis, Consolidation-to-tumor ratio, Solid-predominant, Selective lymph node dissection. (J Thorac Oncol. 2015;10: 930 936) Ever since Cahan 1 first introduced lymph node dissection (LND) of the pulmonary hilum and mediastinum combined with lobectomy (radical lobectomy) as an alternative to pneumonectomy in 1960 and Ramsey et al. 2 subsequently demonstrated the prognostic impact of hilar and mediastinal LND in 1969, systematic LND has been accepted as a standard mode of surgery for resectable non small-cell lung cancer (NSCLC) for over 50 years all around the world. This procedure has been believed to contribute to survival and/or to the determination of more accurate cancer staging. Recently, two controversies have arisen about the concept of LND, which consist of how to perform lymph node harvesting and how to decide the extent of LND. As regards the first issue, some randomized controlled trials that compared systematic LND with systematic lymph node sampling have been performed, and the clinical implications of systematic LND have been discussed in these studies. 3 6 In a recent pivotal study, ACOSOG Z0030, Darling et al. demonstrated that mediastinal LND does not improve survival in patients with early stage NSCLC; however, the results of their study are not generalizable to patients with higher stage tumors. Eventually, they recommend that all patients with resectable NSCLC undergo mediastinal LND because the procedure does not increase mortality or morbidity. 3 As for the other controversy, the extent of LND, some surgeons have retrospectively analyzed the distribution and prevalence of mediastinal lymph node metastasis according to the location of the primary tumor (lobe); they suggested the concept of lobespecific selective LND based on their analyses. 7 12 However, it remains unclear whether selective LND can contribute to a decrease of the incidence of postoperative complications without a negative effect on patient prognosis. 8 It is also a fact that no large-scale prospective randomized controlled trials that compare the clinical implications of systematic LND to lobe-specific selective LND have been performed. 930 Journal of Thoracic Oncology Volume 10, Number 6, June 2015

Journal of Thoracic Oncology Volume 10, Number 6, June 2015 Mediastinal Lymph Node Metastatis in NSCLC Recent developments of radiological examinations have been able to bring more accurate information about the biological malignancy of primary tumors in NSCLC. Especially in cases with small-sized and early NSCLC with ground glass opacity (GGO), preoperative thin-section computed tomography (TSCT) is an important examination for the evaluation and prediction of tumor malignancy. Although there are some reports on the association between radiological findings and histological grade of the tumor in small-sized NSCLC, 13 16 there are none on the relation between radiological findings of the primary tumor on preoperative TSCT and the lobe-specific distribution of mediastinal lymph node metastasis. The aim of this study is to elucidate the optimal candidate for lobe-specific selective LND that reduces the extent of mediastinal LND according to the clinical information including radiological evaluation on TSCT and tumor location in clinical stage (c-stage) I NSCLC patients. MATERIALS AND METHODS Study Population Between January 2003 and December 2009, 1923 consecutive patients underwent pulmonary resection for lung cancer in National Cancer Center Hospital East (Kashiwa, Chiba, Japan). Of these, we reviewed 876 patients who were diagnosed as having c-stage I NSCLC according to the 7 th edition Tumor Node Metastasis classification and underwent complete surgical resection by lobectomy or greater and systematic LND in this study. Complete resection was defined as an absence of residual cancer both macroscopically and microscopically (R0). Patients who underwent limited resections (segmentectomy or wedge resection), incomplete resection were excluded. Histological minorities, such as large cell carcinoma, large cell neuroendocrine carcinoma, carcinoids, pleomorphic carcinoma, and mucoepidermoid carcinoma were excluded, and we performed an analysis only for adenocarcinoma and squamous cell carcinoma cases. We reviewed the medical records of each patient for clinicopathological information as follows: age, sex, smoking status (never or ever smoker), preoperative serum carcinoembryonic antigen (CEA) level (cutoff at the normal upper limit of 5.0 ng/ml), tumor laterality (right or left side), tumor location (right upper lobe [RUL], right middle lobe [RML], right lower lobe [RLL], left upper division [LUD], left lingular division [LLD], or left lower lobe [LLL]), radiological findings (as described below), clinical T factor (c-t factor; ct1a to ct2a), pathological T factor (p-t factor; pt1a to pt4), pathological N factor (p-n factor; pn0 to pn3), pathological stage (p-stage; pia to piv), histology, lymphatic permeation (negative or positive), vascular invasion (negative or positive), pleural invasion (negative or positive), and distribution of each mediastinal lymph node metastasis. Data collection and analyses were approved, and the need to obtain written informed consent from each patient was waived by the Institutional Review Board in June 2012. Preoperative Staging All patients with ct1n0m0 lung cancer were clinically staged using a 5- to 10-mm-collimated contrast-enhanced CT scan (Toshiba, Aquillion 16, TSX301A/GB) of the chest and upper abdomen. Primary lesions were also evaluated using 1- to 2-mm-collimation TSCT images. Lymph nodes larger than 1.0 cm in the short axis on chest CT were clinically defined as metastasis-positive. Mediastinoscopy or positron emission tomography (PET) scan was not routinely performed preoperatively during the period of this study. Most patients except for those with early lung cancer underwent a systemic work-up: either contrast-enhanced brain CT or magnetic resonance imaging of the brain, bone scintigraphy, and contrastenhanced whole-abdominal CT scan. 10 If patients underwent PET scan of the whole body, bone scintigraphy and enhanced whole-abdominal CT scan were often skipped. All patients preoperatively underwent an electrocardiogram and respiratory function test for cardiopulmonary work up. Clinical stage during the study period was determined according to the 6 th edition Tumor Node Metastasis classification in the period of this study. For this study, we restaged all cases according to the 7 th edition Tumor Node Metastasis classification. Radiological Evaluations of Primary Tumors We evaluated primary tumors on chest CT within 1 month before the operation, and their size was determined based on the findings of TSCT scan. GGO was defined as a misty increase in lung attenuation that did not obscure the underlying vascular markings. The solid component was defined as an area of increased opacification, which completely obscured underlying vascular markings. 13 For all tumors, we obtained the maximum dimension of the tumor (tumor) and solid component (consolidation) using a lung window level setting from the TSCT scan images, and estimated the consolidation-totumor ratio (C/T ratio) for each tumor. We defined tumors with C/T ratios less than or equal to 0.5 as GGO-predominant tumors, and tumors with C/T ratios greater than 0.5 as solidpredominant tumors. Each lung nodule on preoperative TSCT scans was reviewed by two thoracic surgeons (T. H. and K.A.), who were unaware of the pathological findings and prognosis. Pathological Evaluation During the period of this study, pathological evaluations of all cases were performed by a single pulmonary pathologist (I. G.). All surgical specimens were fixed with 10% formalin and embedded in paraffin. Serial 4-μm sections were stained using hematoxylin-eosin (HE) stain. The Victoria blue-van Gieson (VVG) method was used to visualize elastic fibers, and the Alcian blue-periodic acid Schiff method was used to detect cytoplasmic mucin production. Blood and lymphatic vessels were identified by HE and VVG staining. Vascular invasion and lymphatic permeation were histologically diagnosed by identifying cancer cells within blood and lymphatic vessels, respectively. Pleural invasion was evaluated by VVG staining. Pleural invasion was considered to be present when tumor cells were identified beyond the elastic layer of the visceral pleura, regardless of tumor exposure on the pleural surface. We evaluated all dissected hilar and mediastinal lymph nodes which harvested in the operation, and each lymph node was evaluated on the maximum cut surface. Histologic analysis of lymph node metastasis was carried out by HE staining. Histological classification of the resected specimens 931

Haruki et al. Journal of Thoracic Oncology Volume 10, Number 6, June 2015 was determined according to the World Health Organization International Histologic Classification of Tumors. 17 Genetic alterations responsible for lung cancer (e.g., EGFR or ALK) were not routinely analyzed during the period of this study. Distribution of Mediastinal Lymph Node Metastasis We categorized and defined mediastinal lymph nodes according to the 7 th edition Tumor Node Metastasis Classification. 18 The sites of mediastinal lymph nodes were grouped based on the IASLC staging 19 as follows: upper mediastinal lymph nodes () included #2R and #4R lymph nodes in cases of right lung cancer. included #4L, #5, and #6 lymph nodes in cases of left lung cancer. In both sides, subcarinal lymph nodes () were the #7 lymph node and lower mediastinal lymph nodes () included the #8 and #9 lymph nodes. Skip metastasis was generally defined as mediastinal lymph node metastasis without intrapulmonary, interlobar, and hilar lymph node metastasis. In this study, mediastinal skip metastasis was defined as metastasis without metastasis for right upper-lobe and left upper-division primary NSCLC and metastasis without metastasis for bilateral lower-lobe primary NSCLC on the basis of lymph node grouping of General Rule for Clinical and Pathological Record of Lung Cancer of the Japan Lung Cancer Society. We analyzed the association of mediastinal lymph node metastasis with clinical factors: age, sex, smoking status, preoperative serum CEA level, tumor laterality, tumor size, and radiological findings. Statistical Analysis Two-category comparison was performed by Pearson s χ 2 test and Fisher s exact test for quantitative data, and continuous data were analyzed by Mann Whitney U test. A p value less than 0.05 was considered to be significant in statistical analyses. Data were analyzed using StatView for Windows version 5.0 (SAS Inc., Cary, NC). RESULTS Clinicopathological Characteristics of Patients The patients included 490 men and 386 women, with a median age of 66 years old (range: 35 89). There were 347 never smokers and 529 ever smokers. The tumor location was RUL in 302 patients, RML in 59 patients, RLL in 187 patients, left upper lobe in 188 patients (LUD in 153 patients and LLD in 35 patients), and LLL in 140 patients. The radiological findings were GGO predominant in 134 patients and solid predominant in 742 patients. The median tumor diameter was 2.5 cm (range: 0.5 5.0). Histologic type was adenocarcinoma in 744 patients, squamous cell carcinoma in 132 patients. Distribution and Incidence of Mediastinal Lymph Node Metastasis Based on the Tumor Location Figure 1 shows the distribution and incidence of lymph node metastasis in each mediastinal region based on the primary tumor location. RUL (n = 302): Twenty-six cases (8.6%) were positive for mediastinal lymph node metastasis. All 26 cases were positive for metastasis, and four cases (1.3%) were positive also for (Fig. 1A). LUD (n = 153): Nine cases (5.8%) were positive for mediastinal lymph node metastasis. Of these, eight cases (5.2%) were positive for metastasis, and one case (0.7%) was positive for (Fig. 1B). RML (n = 59): Five cases (8.5%) were positive for mediastinal lymph node metastasis. Of these, four cases (6.8%) were positive for metastasis, two cases (3.4%) were positive for, and one case (1.7%) was positive for (Fig. 1C). LLD (n = 35): Four cases (11.4%) were positive for mediastinal lymph node metastasis, including a pn3 case that had beaded metastasis of subcarinal and contralateral hilar lymph nodes. Of these, three cases (8.6%) were positive for metastasis, and one case (2.9%) was positive for (Fig. 1D). RLL (n = 187): Thirteen cases (7.0%) were positive for mediastinal lymph node metastasis. Of these, five cases (2.7%) were positive for metastasis, 10 cases (5.3%) were positive for, and six cases (3.2%) were positive for (Fig. 1E). LLL (n = 140): Twenty-three cases (16.4%) were positive for mediastinal lymph node metastasis. Of these, 11 cases (7.8%) were positive for metastasis, 14 cases (10.0%) were positive for, and four cases (2.9%) were positive for (Fig. 1F). Incidence of Mediastinal Lymph Node Metastasis According to Histological Types There were 744 adenocarcinoma cases and 132 squamous cell carcinoma cases in this study. The incidences of mediastinal lymph node metastasis were 9.9% (74 of 744) in adenocarcinoma cases and 4.5% (6 of 132) in squamous cell carcinoma cases, respectively. With regard to six squamous cell carcinoma cases with mediastinal lymph node metastasis, there were five men and a woman. All patients were ever smokers, and had a radiologically solid-predominant tumor. The primary tumor location was RUL in two patients, RLL in one patient, and LLL in three patients. Among them, two upper-lobe cases were positive for metastasis, and three lower-lobe cases were positive for or metastasis, and only one LLL case was positive for metastasis. Associated Clinical Factors of Mediastinal Lymph Node Metastasis in Adenocarcinoma Cases Table 1 shows the associated clinical factors for mediastinal lymph node metastasis in adenocarcinoma cases. Age (p = 0.019), elevated serum CEA level ( 5.0; p < 0.001), and radiological solid-predominant tumors (p < 0.001) showed significant associations with mediastinal lymph node metastasis. Remarkably, there were no mediastinal lymph node metastases in GGO-predominant tumors. Larger tumor size (>2.0 cm, 5.0 cm) was not associated with mediastinal lymph node metastasis (p = 0.158). 932

Journal of Thoracic Oncology Volume 10, Number 6, June 2015 Mediastinal Lymph Node Metastatis in NSCLC A B C RUL N = 302 N = 26 (8.6%) N = 8 (5.2%) LUD N = 153 N = 4 (6.8%) N = 4 (1.3%) N = 0 (0%) N = 1 (0.7%) N = 0 (0%) RML N = 59 N = 2 (3.4%) N = 1 (1.7%) D E F N = 3 (8.6%) N = 5 (2.7%) N = 11 (7.8%) N = 1 (2.9%) N = 0 (0%) LLD N = 35 RLL N = 187 N = 10 (5.3%) N = 6 (3.2%) N = 14 (10.0%) N = 4 (2.9%) LLL N = 140 FIGURE 1. Distribution and incidence of lymph node metastasis in each mediastinal region based on the tumor location. A, Right upper lobe. B, Left upper division. C, Right middle lobe. D, Left lingular division. E, Right lower lobe. F, Left lower lobe., upper mediastinal lymph nodes;, subcarinal lymph nodes;, lower mediastinal lymph nodes. Distributions and Incidences of Mediastinal Skip Metastasis in Adenocarcinoma Cases Table 2 shows a detailed analysis of the distributions and incidences of mediastinal skip metastasis. In 398 RUL and LUD primary lung adenocarcinoma, a total of 33 cases (8.3%) were positive for mediastinal lymph node metastasis, and only one case (0.3%) was negative for metastasis but positive for metastasis. On the other hand, in 257 RLL and LLL primary lung adenocarcinoma, a total of 32 cases (12.5%) were positive for mediastinal lymph node metastasis, and seven cases (2.7%) were negative for metastasis but positive for metastasis. Associated Clinical Factors of Metastasis in Bilateral Lower-Lobe Primary Lung Adenocarcinoma with Radiologically Solid-Predominant Tumor Table 3 shows the association of metastasis with clinical factors including histological type in bilateral lower-lobe primary adenocarcinoma with radiologically solid-predominant tumor. Among these factors, elevated serum CEA level ( 5.0; p < 0.001) showed significant associations with metastasis. DISCUSSION Lobectomy with systematic LND is a globally accepted procedure for resectable primary NSCLC as a standard operation. 1 Although there is a global consensus on the definition of systematic LND that indicates the extent of mediastinal lymph node to dissect and remove, there are some different interpretations of systematic LND among institutions and countries. The European Society of Thoracic Surgeons recommends the dissection of at least three mediastinal nodal regions (but always subcarinal) as a minimum requirement of systematic LND. 20,21 In Japan, mediastinal lymph nodes were divided into three groups (group 2a-1, group 2a-2, and group 2b) according to the primary tumor location, and the extent of LND was determined on the basis of these groups (ND2a- 1, ND2a-2, and ND2b) in the General Rules for Clinical and Pathological Records of Lung Cancer of the Japan Lung Cancer Society 2010. 22 LND is performed based on these rules for almost all cases. Lobe-specific selective LND is conventionally described as ND2a-1 in Japan, which means that the mediastinal tissue containing specific lymph node stations is excised depending on the lobar location of the primary tumor. However, some retrospective studies have revealed the distributions of mediastinal lymph node metastasis in relation to primary tumor location and the significance of lobe-specific selective LND, 7 12,23 25 but its validity or the prognostic effect of lobe-specific selective LND has not been confirmed in a large-scale prospective study. We analyzed the radiological findings of primary tumor on preoperative TSCT and evaluated the association between C/T ratio and mediastinal lymph node metastasis in this study. 933

Haruki et al. Journal of Thoracic Oncology Volume 10, Number 6, June 2015 TABLE 1. Association of Clinicopathological Factors with Mediastinal Lymph Node Metastasis in C-Stage I Lung Adenocarcinoma Total Mediastinal Lymph Node Metastasis N = 744 N = 74 (%) p Value Age (years old) Median 66 (range: 35 89) 0.019 a Sex Men 371 41 (11.1) 0.329 Women 373 33 (8.8) Smoking status Never smoker 345 30 (8.7) 0.326 Ever smoker 399 44 (11.0) Preoperative serum CEA level (ng/ml) <5.0 558 41 (7.3) <0.001 5.0 186 33 (17.7) Tumor laterality Right 478 41 (8.6) 0.098 Left 266 33 (12.4) Tumor size (cm) 2.0 260 20 (7.7) 0.158 >2.0, 5.0 484 54 (11.2) Radiographic findings GGO-predominant 133 0 (0) <0.001 Solid-predominant 611 74 (10.0) a Mann Whitney U test. CEA, carcinoembryonic antigen; GGO, ground glass opacity. TABLE 2. Distributions and Incidences of Lymph Node Metastasis in Each Mediastinal Region Based on the Primary Tumor Location in C-Stage I Lung Adenocarcinoma RUL and LUD primary tumor N = 398 metastasis RLL and LLL primary tumor N = 257 metastasis Metastasis 4 (1.0) 1 (0.3) 28 (7.0) 365 (91.7) Metastasis 8 (3.1) 14 (5.4) 7 (2.7) 225 (87.5) RUL, right upper lobe; LUD, left upper division; RLL, right lower lobe; LLL, left lower lobe;, upper mediastinal lymph nodes;, subcarinal lymph nodes. TABLE 3. Association of Clinicopathological Factors with Upper Mediastinal Lymph Node Metastasis in Bilateral Lower- Lobe Primary Lung Adenocarcinoma with Radiologically Solid-Predominant Tumor Total Upper Mediastinal Lymph Node Metastasis N = 200 N = 15 (%) p Value Age (years old) Median 67 (range: 37 89) 0.055 a Sex Men 98 7 (7.1) >0.999 Women 102 8 (7.8) Smoking status Never smoker 95 9 (9.5) 0.422 Ever smoker 105 6 (5.7) Preoperative serum CEA level (ng/ml) <5.0 144 4 (2.8) <0.001 5.0 56 11 (19.6) Tumor laterality Right 114 5 (4.4) 0.062 Left 86 10 (11.6) Tumor size (cm) 2.0 53 2 (3.8) 0.362 >2.0, 5.0 147 13 (8.8) Tumor location Upper segment 68 8 (11.8) 0.154 Basal segment 132 7 (5.3) a Mann Whitney U test. CEA, carcinoembryonic antigen. In previous reports, the incidence of mediastinal lymph node metastasis was extremely low in c-stage I NSCLC cases with GGO-predominant tumor. 13,16,26 In recent years, the opportunity to detect lung nodules with a GGO component has been increasing owing to the wide spread of CT screening and the resolution-enhancement technology of TSCT. 27 29 Several authors have reported that it was useful to evaluate the detailed radiological features to predict tumor invasiveness. 14,16 A prospective radiological study in Japan (JCOG0201) has been conducted to demonstrate the predictive TSCT criterion of pathological noninvasiveness for c-stage I NSCLC. 15 In addition, there are a few reports that describe detailed analyses based on radiological findings on TSCT for the distribution and prevalence of mediastinal lymph node metastasis in c-stage I NSCLC. Our results suggest that selective LND might be acceptable for GGO-predominant cases in the sense that there would probably be no remaining metastatic lymph nodes by selective LND. Moreover, it might be unnecessary to perform hilar and mediastinal LND itself for GGO-predominant lung cancer. On the other hand, there were a lot of cases with unexpected hilar and mediastinal lymph node metastasis in solid-predominant tumors. All primary tumors with hilar and mediastinal lymph 934

Journal of Thoracic Oncology Volume 10, Number 6, June 2015 Mediastinal Lymph Node Metastatis in NSCLC node metastasis were solid predominant, so we cannot help but be careful to apply a reduction in the extent of LND to these populations. From the perspective of lobe-specific selective LND, the incidence of metastasis was only 1.3% (5 of 398 cases), and only one case (0.3%) was positive for metastasis but negative for metastasis in primary adenocarcinoma located in the RUL and LUD. There were no predictive clinical factors of metastasis in RUL and LUD primary adenocarcinoma; however, there would be some possibility to omit dissection because the incidence of metastasis was extremely low, even in solidpredominant c-stage I NSCLC cases. On the other hand, the incidence of metastasis in lower-lobe primary NSCLC was much higher than that of metastasis in upper-lobe primary NSCLC because of its lymphatic drainage pathway. With regard to the selective LND of lower-lobe NSCLC, there are some reports that demonstrate the usefulness of intraoperative frozen section diagnosis of mediastinal lymph nodes 9,30 ; namely, dissection might be unnecessary for lower-lobe tumors when the is intraoperatively negative. However, half of lower-lobe cases with metastasis had no metastasis of in our study. A lot of care must be taken for the application of selective LND for solid- predominant tumors located in bilateral lower lobes. According to the analysis in our study, CEA elevation was a significant associated factor of metastasis in lower-lobe primary lung adenocarcinoma with radiologically solid-predominant tumor. Hattori et al. 26 revealed that CEA elevation was a significant predictor of lymph node involvement in c-stage I NSCLC cases, and they suggested that incomplete dissection or sampling of lymph nodes for solid-predominant tumor could result in locoregional recurrence. Considering the results of our study, it might be possible to omit dissection for lower-lobe primary tumors with a normal level of CEA; however, it is necessary to investigate more cases with long term follow-up to confirm the possibility. We should also pay careful attention to the possibility of skip mediastinal lymph node metastasis. There were one right upper-lobe case with skip metastasis and seven lowerlobe cases with skip metastasis in our study. It is well known that there are pulmonary lymphatic pathways directly to the mediastinum. Gorai et al. 31 suggested that the incidence of skip N2 was significantly associated with pleural involvement in patients with clinical IA NSCLC and in those with nodal metastasis, and assumed an involvement of the subpleural lymphatic drainage. In addition, Takahashi et al. 32,33 demonstrated a predictable pattern of lobar lymphatic drainage to the mediastinum by the assessment of CT studies of patients with primary complex histoplasmosis consisting of a single lobe pulmonary lesion and mediastinal nodal disease. They revealed that there were seven of 122 (5.7%) upper-lobe cases with skip involvement and 14 of 129 (10.8%) lowerlobe cases with skip involvement in their study. The data from their study would lead to our good understanding of lymphatic drainage pathways of each lobe and segment to the mediastinum. Limitations of this study include the potential introduction of surgical and selection bias associated with the retrospective assessment of data. Another limitation of this study is that the presented data are derived from a single institution, and there were not a sufficient number of cases, especially nonadenocarcinoma cases, to reach a definitive conclusion. In addition, the calculation of C/T ratio was performed only by thoracic surgeons, not including thoracic radiologist. There was no internal observer who decided whether there were any interreader variabilities. Finally, preoperative mediastinal evaluations by PET/CT or endobronchial ultrasonography-transbronchial needle aspiration were not routinely performed during the period of this study because of the inclusion of almost data from old cases, collected at a time when these examinations were less established. It is imperative to perform a large-scale prospective randomized trial that compares systematic LND to selective LND for solid-predominant NSCLC. In conclusion, we demonstrated the association of clinical factors including radiological findings on TSCT with mediastinal lymph node metastasis. It would be acceptable to perform selective LND for c-stage I GGO-predominant tumors; moreover, it might be possible to omit mediastinal LND itself for those cases. On the other hand, we should be careful to apply selective LND to c-stage I solid-predominant tumors, especially those located in the lower lobe. CEA elevation was significantly associated with metastasis in lower-lobe primary c-stage I lung adenocarcinoma with radiologically solid-predominant tumors. It would be necessary to perform well-designed prospective clinical trials to evaluate the indication of selective LND. REFERENCES 1. Cahan WG. Radical lobectomy. J Thorac Cardiovasc Surg 1960;39:555 572. 2. Ramsey HE, Cahan WG, Beattie EJ, Humphrey C. The importance of radical lobectomy in lung cancer. J Thorac Cardiovasc Surg 1969;58:225 230. 3. Darling GE, Allen MS, Decker PA, et al. Randomized trial of mediastinal lymph node sampling versus complete lymphadenectomy during pulmonary resection in the patient with N0 or N1 (less than hilar) non-small cell carcinoma: Results of the American College of Surgery Oncology Group Z0030 Trial. J Thorac Cardiovasc Surg 2011;141:662 670. 4. Wu N, Yan S, Lv C, et al. Comparison of systematic mediastinal lymph node dissection versus systematic sampling for lung cancer staging and completeness of surgery. J Surg Res 2011;171:e169 e173. 5. Izbicki JR, Passlick B, Pantel K, et al. Effectiveness of radical systematic mediastinal lymphadenectomy in patients with resectable non-small cell lung cancer: Results of a prospective randomized trial. Ann Surg 1998;227:138 144. 6. Sugi K, Nawata K, Fujita N, et al. Systematic lymph node dissection for clinically diagnosed peripheral non-small-cell lung cancer less than 2 cm in diameter. World J Surg 1998;22:290 4; discussion 294. 7. Asamura H, Nakayama H, Kondo H, Tsuchiya R, Naruke T. Lobe-specific extent of systematic lymph node dissection for non-small cell lung carcinomas according to a retrospective study of metastasis and prognosis. J Thorac Cardiovasc Surg 1999;117:1102 1111. 8. Okada M, Sakamoto T, Yuki T, Mimura T, Miyoshi K, Tsubota N. Selective mediastinal lymphadenectomy for clinico-surgical stage I nonsmall cell lung cancer. Ann Thorac Surg 2006;81:1028 1032. 9. Watanabe S, Asamura H, Suzuki K, Tsuchiya R. The new strategy of selective nodal dissection for lung cancer based on segment-specific patterns of nodal spread. Interact Cardiovasc Thorac Surg 2005;4:106 109. 10. Aokage K, Yoshida J, Ishii G, Hishida T, Nishimura M, Nagai K. Subcarinal lymph node in upper lobe non-small cell lung cancer patients: Is selective lymph node dissection valid? Lung Cancer 2010;70:163 167. 935

Haruki et al. Journal of Thoracic Oncology Volume 10, Number 6, June 2015 11. Ishiguro F, Matsuo K, Fukui T, Mori S, Hatooka S, Mitsudomi T. Effect of selective lymph node dissection based on patterns of lobe-specific lymph node metastases on patient outcome in patients with resectable non-small cell lung cancer: A large-scale retrospective cohort study applying a propensity score. J Thorac Cardiovasc Surg 2010;139:1001 1006. 12. Cerfolio RJ, Bryant AS. Distribution and likelihood of lymph node metastasis based on the lobar location of nonsmall-cell lung cancer. Ann Thorac Surg 2006;81:1969 73; discussion 1973. 13. Suzuki K, Asamura H, Kusumoto M, Kondo H, Tsuchiya R. Early peripheral lung cancer: Prognostic significance of ground glass opacity on thin-section computed tomographic scan. Ann Thorac Surg 2002;74:1635 1639. 14. Ohde Y, Nagai K, Yoshida J, et al. The proportion of consolidation to ground-glass opacity on high resolution CT is a good predictor for distinguishing the population of non-invasive peripheral adenocarcinoma. Lung Cancer 2003;42:303 310. 15. Suzuki K, Koike T, Asakawa T, et al.; Japan Lung Cancer Surgical Study Group (JCOG LCSSG). 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;6:751 756. 16. Suzuki K, Kusumoto M, Watanabe S, Tsuchiya R, Asamura H. Radiologic classification of small adenocarcinoma of the lung: Radiologic-pathologic correlation and its prognostic impact. Ann Thorac Surg 2006;81:413 419. 17. Travis WD, Branbilla E, Muller-Hermelink HK, Harris CC. World Health Organization Classification of Tumors. Pathology and Genetics of Tumors of the Lung, Pleura, Thymus and Heart. Lyon: IARC Press, 2004. 18. Goldstraw P, Crowley J, Chansky K, et al.; International Association for the Study of Lung Cancer International Staging Committee; Participating Institutions. 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 tumours. J Thorac Oncol 2007;2:706 714. 19. Rusch VW, Asamura H, Watanabe H, Giroux DJ, Rami-Porta R, Goldstraw P; Members of IASLC Staging Committee. The IASLC lung cancer staging project: A proposal for a new international lymph node map in the forthcoming seventh edition of the TNM classification for lung cancer. J Thorac Oncol 2009;4:568 577. 20. Rami-Porta R, Wittekind C, Goldstraw P; International Association for the Study of Lung Cancer (IASLC) Staging Committee. Complete resection in lung cancer surgery: Proposed definition. Lung Cancer 2005;49:25 33. 21. Lardinois D, De Leyn P, Van Schil P, et al. ESTS guidelines for intraoperative lymph node staging in non-small cell lung cancer. Eur J Cardiothorac Surg 2006;30:787 792. 22. The Japan Lung Cancer Society. General rule for clinical and pathological record of lung cancer. Japan Lung Cancer Society. Nihon Rhinso 2010; 47 61. 23. Shimada Y, Saji H, Kakihana M, et al. Retrospective analysis of nodal spread patterns according to tumor location in pathological N2 non-small cell lung cancer. World J Surg 2012;5:5. 24. Uehara H, Sakao Y, Mun M, et al. Prognostic value and significance of subcarinal and superior mediastinal lymph node metastasis in lower lobe tumours. Eur J Cardiothorac Surg 2010;38:498 502. 25 Riquet M, Rivera C, Pricopi C, et al. Is the lymphatic drainage of lung cancer lobe specific? A surgical appraisal. Eur J Cardiothorac Surg. 2015;47:543 549. 26. Hattori A, Suzuki K, Matsunaga T, et al. Is limited resection appropriate for radiologically solid tumors in small lung cancers? Ann Thorac Surg 2012;94:212 215. 27. Flieder DB, Vazquez M, Carter D, et al. Pathologic findings of lung tumors diagnosed on baseline CT screening. Am J Surg Pathol 2006;30:606 613. 28. Sone S, Nakayama T, Honda T, et al. Long-term follow-up study of a population-based 1996-1998 mass screening programme for lung cancer using mobile low-dose spiral computed tomography. Lung Cancer 2007;58:329 341. 29. Fujikawa A, Takiguchi Y, Mizuno S, et al. Lung cancer screening comparison of computed tomography and X-ray. Lung Cancer 2008;61:195 201. 30. Okada M, Tsubota N, Yoshimura M, Miyamoto Y. Proposal for reasonable mediastinal lymphadenectomy in bronchogenic carcinomas: Role of subcarinal nodes in selective dissection. J Thorac Cardiovasc Surg 1998;116:949 953. 31. Gorai A, Sakao Y, Kuroda H, et al. The clinicopathological features associated with skip N2 metastasis in patients with clinical stage IA nonsmall-cell lung cancer. Eur J Cardiothorac Surg 2015;47:653 658. 32. Takahashi K, Stanford W, Van Beek E, Thompson B, Mullan B, Sato Y. Mediastinal lymphatic drainage from pulmonary lobe based on CT observations of histoplasmosis: Implications for minimal N2 disease of nonsmall-cell lung cancer. Radiat Med 2007;25:393 401. 33. Takahashi K, Sasaki T, Nabaa B, van Beek EJ, Stanford W, Aburano T. Pulmonary lymphatic drainage to the mediastinum based on computed tomographic observations of the primary complex of pulmonary histoplasmosis. Acta Radiol 2012;53:161 167. 936