Intraoperative diagnosis of lymph node metastasis in non-small-cell lung cancer by a semi-dry dot-blot method

Transcription

1 European Journal of Cardio-Thoracic Surgery Advance Access published April 7, 2015 European Journal of Cardio-Thoracic Surgery (2015) 1 7 doi: /ejcts/ezv118 ORIGINAL ARTICLE Cite this article as: Tomoshige K, Tsuchiya T, Otsubo R, Oikawa M, Yamasaki N, Matsumoto K et al. Intraoperative diagnosis of lymph node metastasis in non-small-cell lung cancer by a semi-dry dot-blot method. Eur J Cardiothorac Surg 2015; doi: /ejcts/ezv118. Intraoperative diagnosis of lymph node metastasis in non-small-cell lung cancer by a semi-dry dot-blot method Koichi Tomoshige a, Tomoshi Tsuchiya a, Ryota Otsubo a, Masahiro Oikawa a, Naoya Yamasaki a, Keitaro Matsumoto a, Takuro Miyazaki a, Tomayoshi Hayashi b, Naoe Kinoshita b, Atsushi Nanashima a and Takeshi Nagayasu a, * a Department of Surgical Oncology, Nagasaki University Graduate School of Biomedical Sciences, Nagasaki, Japan b Department of Pathology, Nagasaki University Hospital, Nagasaki, Japan * Corresponding author. Department of Surgical Oncology, Nagasaki University Graduate School of Biomedical Science, Sakamoto, Nagasaki , Japan. Tel: ; fax: ; nagayasu@nagasaki-u.ac.jp (T. Nagayasu). Received 7 December 2014; received in revised form 12 February 2015; accepted 24 February 2015 Abstract OBJECTIVES: Sublobar resection procedures, such as segmentectomy and wedge resection, can be used for resectable lung cancer when the cancer is small or the condition of the patient is poor. In such cases, intraoperative lymph node (LN) exploration is necessary to avoid incomplete resection of potential N1 or N2 disease. The semi-dry dot-blotting (SDB) method was developed to detect intraoperative LN metastasis as a quick, cost-effective procedure that does not require special technical expertise. This study examined whether SDB can sufficiently identify LN metastasis in lung cancer patients. METHODS: This study prospectively examined 147 LNs from 50 lung cancer patients who underwent surgery at Nagasaki University Hospital between April 2011 and June The SDB method uses antigen antibody reactions with anti-pancytokeratin as the primary antibody and detects cancer cells using chromogen. To identify LN metastases, each LN was examined by the SDB method during surgery along with intraoperative pathological diagnosis (ope-dx) and permanent pathological diagnosis ( permanent-dx). RESULTS: Compared with permanent-dx, SDB offered 94.7% sensitivity, 97.7% specificity and 97.2% accuracy, while ope-dx exhibited 84.2% sensitivity, 100% specificity and 98.0% accuracy. For 3 cases, micrometastases were detected by the SDB method but not by ope-dx. Three LNs from lobar stations showed pseudo-positive results by the SDB method because of the presence of alveolar epithelium. CONCLUSIONS: The SDB method offers acceptably high accuracy in detecting LN metastasis, especially for mediastinal LNs, and represents a potential alternative for the intraoperative diagnosis of LN metastasis, even in the absence of a pathologist. Keywords: Semi-dry dot-blot Lung cancer Lymph node Metastasis Cytokeratin INTRODUCTION Sublobar resection procedures, such as segmentectomy and wedge resection, have recently been recognized as treatment options for early, small non-small-cell lung cancer (NSCLC), although lobectomy remains the standard treatment procedure for NSCLC patients [1 5]. In NSCLC cases treated by sublobar resection, intraoperative lymph node (LN) exploration is necessary to avoid incomplete resection, as N1 or N2 disease may be present; % of clinical stage I patients have mediastinal LN involvement [6, 7]. Intraoperative evaluation of LNs is ordinarily performed by histological examination using frozen sections. While the cut surface Presented at the 28th Annual Meeting of the European Association for Cardio- Thoracic Surgery, Milan, Italy, October alone is usually used for rapid diagnosis, this approach requires equipment for making frozen sections and requires an on-call pathologist. In addition, the metastasis detection rate of this method is insufficient due to the difficulty of detecting metastases by examining frozen sections [8]. The semi-dry dot-blotting (SDB) method is reportedly a useful diagnostic tool for sentinel LN metastasis in breast cancer [9]. Using dot-blot analysis techniques, the SDB method visualizes the presence of cancer cells by washing sectioned LNs with antipancytokeratin antibody (AE1/AE3) and chromogen on a dot-blot membrane. The sensitivity of this method is extremely high, detecting as little as 0.01 mg/ml of protein extracted from cancer tissue or 20 cells from a tissue suspension [10]. In addition, the method is highly cost-effective because it requires simple materials, including cutting blades, primary and secondary antibodies, chromogen, filters, absorption papers and dot-blot membranes. The Author Published by Oxford University Press on behalf of the European Association for Cardio-Thoracic Surgery. All rights reserved.

2 2 K. Tomoshige et al. / European Journal of Cardio-Thoracic Surgery The detection time for this method is 40 min, which is similar to the time required for intraoperative pathological diagnosis [9]. The present study prospectively assessed whether the SDB method could adequately diagnose LN metastasis from NSCLC. We compared the sensitivity, specificity and accuracy of the SDB method with those of intraoperative evaluation and evaluated the usefulness of this method in the field of lung cancer, particularly for intraoperative diagnosis. MATERIALS AND METHODS Study patients Patients who were scheduled to undergo operations for NSCLC were enrolled in this study from April 2011 to June This study was conducted with the approval of the Genetic and Medical Ethics Commission at Nagasaki University (Approval No ), and all subjects signed their informed consent prior to participating in the study. Semi-dry dot-blot method For each case, the single or the biggest LN from each site was considered the representative LN of that site and was examined by the SDB method. Each LN was trimmed of surplus adipose tissue and cut along the maximum diameter by a surgeon. Divided LNs were placed in 15 ml silicon tubes and washed for 30 s with 2 ml of phosphate-buffered saline in the operation theatre, and the lavage fluid was used for the diagnosis of LN metastasis by the SDB method. This process was simple and only a few minutes of training was required before performing the examination. An outline of the procedure for the SDB method is shown in Fig. 1. Washed LNs were delivered to the pathological division for intraoperative pathological examination. The lavage fluid was delivered to the laboratory room and research assistants centrifuged it to collect suspended cells as a pellet. After the pellet was lysed with lysis buffer, filtration through a spin column filter was performed to remove connective tissues and adipose tissues. For preparation of the absorption paper, a dot-blot membrane ( polyethersulfone membrane) was soaked in the wash buffer. Anti-pancytokeratin (AE1/AE3; Nichirei, Tokyo, Japan) was used as the primary antibody, and peroxidase-coupled anti-mouse immunoglobulin G (ImmPRESS anti-mouse IgG; Vector Laboratories, Burlingame, CA, USA) was used as the secondary antibody. Two circles for each sample were drawn on the membrane. Both primary and secondary antibodies were placed onto one circle for the actual examination, and primary antibody alone was placed onto the other circle as a reference sample. The membrane was incubated with the two antibodies for 3 min at room temperature. The reaction was visualized using the VECTOR VIP substrate kit (Vector Laboratories), which produces a purple chromogen in the presence of peroxidase enzyme. A positive result was defined as a denser appearance of chromogen in the circle of the actual examination than in the reference circle. A negative result was defined as an equal density of chromogen in both circles. Pathological examination All LNs examined by the SDB method were examined intraoperatively and the final diagnosis was achieved using formalin-fixed Figure 1: Outline of the SDB method. Each lymph node (LN) is cut at the maximum diameter and washed with phosphate-buffered saline. Cell components are centrifuged and collected as a pellet. The pellet is lysed with lysis buffer to extract protein. The prepared specimen is spotted onto a dot-blot membrane, and semi-dry conditions are created by absorption paper transfused with washed buffer. The membrane is incubated with anti-pancytokeratin antibody (AE1/AE3) and secondary antibody, and the reaction is visualized using a chromogen on a dot-blot membrane.

3 K. Tomoshige et al. / European Journal of Cardio-Thoracic Surgery 3 and paraffin-embedded sections with haematoxylin and eosin staining. The analyses of the SDB method, intraoperative pathology and permanent pathology results were performed independently. The Tumor, Node, Metastasis (TNM) staging was conducted according to the seventh edition of the International Association for the Study of Lung Cancer [11]. RESULTS In this study, 147 LNs were obtained from 50 lung cancer patients who underwent surgery at Nagasaki University Hospital from April 2011 to June The patient characteristics are presented in Table 1. Forty-nine patients underwent lobectomy and mediastinal LN dissection and 1 patient underwent mediastinal LN sampling for TNM staging. Validation of semi-dry dot-blot diagnosis of LN metastasis compared with permanent pathological diagnosis Of the 147 LNs eligible for analysis, 19 (12.9%) were positive and 128 (87.1%) were negative for malignancy by permanent pathological diagnosis ( permanent-dx). By the SDB method, 21 (14.3%) were positive and 126 (85.7%) were negative for malignancy. Thus, the accuracy and sensitivity of the SDB method compared with permanent-dx were 97.3% (Table 2) and 94.7%, respectively. One of the 19 positive LNs was not detected by the SDB method. The specificity of the SDB method compared with permanent-dx was 97.7%, and 3 (3.1%) LNs determined to be pathologically negative by permanent-dx were identified as positive by the SDB method. The 1 false-negative and 3 false-positive LNs identified by the SDB method were subjected to further analysis. Validation of frozen section-based intraoperative pathological diagnosis of lymph node metastasis compared with permanent pathological diagnosis Of the 147 LNs eligible for analysis, 16 (12.2%) were positive and 131 (89.1%) were negative for malignancy by intraoperative pathological diagnosis (ope-dx), resulting in 98.0% accuracy for ope-dx compared with permanent-dx (Table 3). The sensitivity of ope-dx compared with permanent-dx was 84.2%. However, 3 LNs were assessed as negative by ope-dx and positive by permanent-dx. The specificity of ope-dx compared with permanent-dx was 100%. The 3 false-negative LNs identified by ope-dx were subjected to further analysis. Pathological analysis of 1 false-positive and 3 false-negative lymph nodes by the semi-dry dot-blot method and 3 false-negative lymph nodes by intraoperative pathological diagnosis We analysed 7 LNs with discrepancies in the results of LN malignancy diagnosis by the SDB method and permanent-dx or between ope-dx and permanent-dx (Table 4). Figure 2A shows the sites and numbers of these LNs. Table 2: Comparison of the SDB method and permanent pathological diagnosis in the detection of cancer cells among clinically dissected LNs Pathology (permanent) Positive Negative Total Table 1: Clinicopathological characteristics of patients SDB Positive Negative total Characteristic Values Patients, n 50 Age, years 67.1 (34 84) Sex, n Male 32 Female 18 Histology, n Adenocarcinoma 35 Squamous cell carcinoma 10 Large cell carcinoma 1 Other histology 3 c-stage, n IA 31 IB 6 IIA 5 IIB 3 IIIA 5 p-stage, n IA 21 IB 11 IIA 5 IIB 4 IIIA 8 IIIB 1 Sensitivity = 18/19 (94.7%). Specificity = 125/128 (97.7%). Accuracy = 143/147 (97.3%). Table 3: Comparison of intraoperative pathological diagnosis and permanent pathological diagnosis in the detection of cancer cells among clinically dissected LNs Pathology (permanent) Positive Negative Total Pathology (intraoperative) Positive Negative Total Sensitivity = 16/19 (84.2%). Specificity = 128/128 (100%). Accuracy = 144/147 (98%).

4 4 K. Tomoshige et al. / European Journal of Cardio-Thoracic Surgery Table 4: Comparison of the 7 cases with discordant results between the SDB method and pathological diagnosis Case Stage Location Permanent pathological diagnosis Intraoperative pathological diagnosis SDB 1 Stage Ib Iobar Negative Negative Positive 2 Stage IIa Iobar Negative Negative Positive 3 Stage Ib Interlobar Negative Negative Positive 4 Stage IIIa Subcarinal Positive Positive Negative 5 Stage IIIb Subcarinal Positive Negative Positive 6 Stage IIIa Paratracheal Positive Negative Positive 7 Stage IIb Interlobar Positive Negative Positive Figure 2: (A) The schematic diagram of the location of lymph nodes which had different diagnoses between the SDB method and permanent pathological diagnosis or between intraoperative pathological diagnosis and permanent pathological diagnosis. The numbers indicate lymph node stations as below; : paratracheal, : subcarinal, : interlobar, : lobar. An immunochemical staining of a false-positive result using the SDB method (asterisk) is shown in (B). LN; lymph node: ope-dx; intraoperative pathological diagnosis. (B) AE1/AE immunohistochemical staining of LN resected from a patient with lung cancer, with a false-positive result from the SDB method (Case 3; Table 4). Alveolar epithelial tissue contiguous with the normal LN was stained, and cytokeratin from alveolar epithelial tissue led to the positive diagnosis with SDB. Bar, 200 µm. Of the 3 false-positive LNs identified by the SDB method, immunohistochemical staining for AE1/AE3, an epithelial marker, revealed that all 3 were contaminated with alveolar epithelial tissue from adjacent normal LN tissues (Fig. 2B). For the false-negative LN, which was from a subcarinal site, thin sections stained with haematoxylin and eosin were shown to contain malignant cells. Immunohistochemical staining of AE1/AE3 for permanent-dx revealed that all 3 false-negative LNs identified by ope-dx were micrometastases (Fig. 3). Validation of the semi-dry dot-blot method in the diagnosis of mediastinal lymph node metastasis in the clinical setting of non-small-cell lung cancer Discrepancies between the results of the SDB method and permanent-dx or between ope-dx and permanent-dx (Fig. 2A) were only observed for LNs from four sites (interlobar, lobar, subcarinal and paratracheal), and the interlobar and lobar LNs that

5 K. Tomoshige et al. / European Journal of Cardio-Thoracic Surgery 5 Figure 3: Haematoxylin and eosin staining and immunohistochemical staining of micrometastatic LN, which was not detected by intraoperative pathological diagnosis (Case 5; Table 4). Representative images of haematoxylin and eosin staining (A) and positive staining for AE1/AE3 (B). Bar, 200 µm. Table 5: Comparison of data between the SDB method and OSNA Permanent pathological diagnosis Diagnostic method Sample (organ) Sensitivity (%) Specificity (%) Accuracy (%) Specimen loss SDB (present study) LN (lung) ( ) SDB (Otsubo et al.[9]) LN (breast) OSNA (Hayama et al.[12]) LN (lung) NA (+) OSNA (Inoue et al.[13]) LN (lung) NA NA 98.8 NA: not available. had inconsistent results were all hilar LNs, while the subcarinal and paratracheal LNs that had consistent results were all mediastinal LNs. A total of 104 mediastinal LNs were examined. Of these, 12 were positive by the SDB method and 13 were positive by permanent-dx; only 1 mediastinal LN was a false-negative by the SDB method (see Supplementary material, Table S1). All 91 negative LNs by the SDB method were confirmed to be negative by permanent-dx. Accordingly, the sensitivity, specificity and accuracy of the SDB method were 92.3, 100 and 99.0%, respectively. DISCUSSION The SDB method is a novel procedure for detecting LN metastasis; it is based on the dot-blot method and was developed by Hirakawa et al.[10]. The SDB method showed high accuracy in the present study, and the sensitivity, specificity and accuracy of the SDB method compared with permanent-dx were 94.7, 97.7 and 97.3%, respectively. Otsubo et al. [9] used the SDB method to test 174 sentinel LNs from 100 breast cancer patients. The sensitivity, specificity and accuracy of the SDB method were not inferior to those of pathological examination and were 100, 98.0 and 98.3%, respectively. The present results are comparable with the previous results in breast cancer patients. Furthermore, the present results revealed that the SDB method is comparable to ope-dx in lung cancer. The sensitivity, specificity and accuracy of ope-dx were 84.2, 100 and 98.0%, respectively. Given that the time required for the SDB method was similar to that for frozen section-based intraoperative pathological examination, 40 min, the SDB method is considered feasible for the ope-dx of LN metastasis [9]. The diagnostic accuracy of the SDB method is also comparable to another novel technique for the diagnosis of LN metastasis: one-step nucleic amplification (OSNA) (Sysmex, kobe, Japan [12 15]. This assay uses cytokeratin 19 mrna as a marker and measures copy numbers of cytokeratin 19 mrna after loop-mediated isothermal amplification. A comparison of the data obtained using the SDB method and OSNA is presented in Table 5. Inthefield of lung cancer, Hayama et al.[12] reported high sensitivity (100%) and specificity (91.7%) for OSNA, and Inoue et al.[13] reported a high positive predictive value (95.0%), a high negative predictive value (99.3%) and high accuracy (98.8%) for OSNA. However, the SDB method offers some advantages over OSNA. First, the SDB method is a simple procedure employing antigen antibody reactions and does not require special equipment. In contrast, OSNA requires equipment for the quantitative reverse transcription polymerase chain reaction and the homogenization of LN tissue. Accordingly, the cost of the SDB method is more than 10 times lower than that of OSNA [9]. Secondly, because only the lavage fluids of LNs are used in the SDB method, the procedure does not cause specimen loss. LN sections can be used for both the SDB method and pathological diagnosis. In contrast, OSNA and pathological examination cannot be performed on the same section because OSNA requires homogenized tissue. Some lung-specific technical issues arise with the SDB method. LNs of the lung often suffer anthracosis, which changes the colour of lavage fluid to grey or black. Because LN metastasis is judged by the purple coloration of the membrane in the SDB method,

6 6 K. Tomoshige et al. / European Journal of Cardio-Thoracic Surgery carbon remaining on the membrane complicates the determination of metastasis. A longer centrifugation ( rpm, 5 min) can avoid the retention of charcoal in the sample. Another important point for sample preparation for the SDB method is that epithelial cell or pleural cell contamination can result in a misdiagnosis, as these cells express cytokeratin. In the present study, detailed pathological examinations revealed that alveolar epithelial tissues were not sufficiently removed from the 3 false-positive LNs. Those LNs were resected from interlobar and lobar areas (Fig. 2A). In contrast, contaminating lung epithelial cells were not present in LNs resected from mediastinal areas, including paratracheal, prevascular, prevertebral, subaortic, paraaortic, subcarinal, paraoesophageal and pulmonary ligament sites, and the specificity and accuracy of our results increased to 100 and 99.0%, respectively, when the data of the interlobar and lobar LNs were excluded from the analysis. The present results indicate that careful removal of adherent lung tissue from LNs in interlobar and lobar areas is crucial for this method. We believe the SDB method is valuable for intraoperative diagnosis of LNs, especially for mediastinal LNs because mediastinal LN dissection is an important part of lung surgery that is associated with improved outcomes [16], and % of clinical stage I patients have mediastinal LN involvement [6, 7]. CONCLUSIONS The SDB method is a novel and effective procedure for the diagnosis of LN metastasis in lung cancer patients. The high accuracy of this method is comparable with that of ope-dx. The advantages of this procedure are its simplicity, high accuracy, cost-effectiveness and preservation of LN tissue for pathological examination. The SDB method will help in the intraoperative diagnosis of LN metastasis in many institutions because technical experts and skilled pathologists are not needed for performing this procedure. SUPPLEMENTARY MATERIAL Supplementary material is available at EJCTS online. ACKNOWLEDGEMENTS We thank Masahiro Nakashima from the Department of Tumor and Diagnostic Pathology at Nagasaki University for kindly advising us regarding our immunohistochemistry experiments. Conflicts of interest: none declared. REFERENCES [1] Schuchert MJ, Kilic A, Pennathur A, Nason KS, Wilson DO, Luketich JD et al. Oncologic outcomes after surgical resection of subcentimeter non-small cell lung cancer. Ann Thorac Surg 2011;91:1681 7; discussion [2] 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: [3] Kates M. Survival following lobectomy and limited resection for the treatment of stage I non-small cell lung cancer 1 cm in size: a review of SEER data. Chest 2011;139: [4] Kilic A, Schuchert MJ, Pettiford BL, Pennathur A, Landreneau JR, Landreneau JP et al. Anatomic segmentectomy for stage I non-small cell lung cancer in the elderly. Ann Thorac Surg 2009;87:1662 6; discussion [5] Asamura H, Suzuki K, Watanabe S, Matsuno Y, Maeshima A, Tsuchiya R. A clinicopathological study of resected subcentimeter lung cancers: a favorable prognosis for ground glass opacity lesions. Ann Thorac Surg 2003;76: [6] Nomori H, Watanabe K, Ohtsuka T, Naruke T, Suemasu K. In vivo identification of sentinel lymph nodes for clinical stage I non-small cell lung cancer for abbreviation of mediastinal lymph node dissection. Lung Cancer 2004;46: [7] Inoue M, Minami M, Sawabata N, Utsumi T, Kadota Y, Shigemura N et al. Clinical outcome of resected solid-type small-sized c-stage IA non-small cell lung cancer. Eur J Cardiothorac Surg 2010;37: [8] Dhaliwal CA, Andrews TD, Walker WS, Wallace WA. Histological evaluation of preoperative mediastinoscopy lymph node biopsies in non-small cell lung cancer. J Clin Pathol 2014;67: [9] Otsubo R, Oikawa M, Hirakawa H, Shibata K, Abe K, Hayashi T et al. Novel diagnostic procedure for determining metastasis to sentinel lymph nodes in breast cancer using a semi-dry dot-blot method. Int J Cancer 2014;134: [10] Hirakawa H, Shibata K, Ohzono E. Use of a semi-dry dot-blot for rapid detection of lymph node metastasis. Clin Chim Acta 2010;411: [11] Rusch VW, Asamura H, Rami-porta R, Goldstraw P, Committee S. The IASLC lung cancer staging project. J Thorac Oncol 2009;4: [12] Hayama M, Chida M, Karube Y, Tamura M, Kobayashi S, Oyaizu T et al. One-step nucleic acid amplification for detection of lymph node metastasis in lung cancer. Ann Thorac Cardiovasc Surg 2014;20: [13] Inoue M, Hiyama K, Nakabayashi K, Morii E, Minami M, Sawabata N et al. An accurate and rapid detection of lymph node metastasis in non-small cell lung cancer patients based on one-step nucleic acid amplification assay. Lung Cancer 2012;78: [14] Visser M, Jiwa M, Horstman A, Brink AATP, Pol RP, van Diest P et al. Intra-operative rapid diagnostic method based on CK19 mrna expression for the detection of lymph node metastases in breast cancer. Int J Cancer 2008;122: [15] Tsujimoto M, Nakabayashi K, Yoshidome K, Kaneko T, Iwase T, Akiyama F et al. One-step nucleic acid amplification for intraoperative detection of lymph node metastasis in breast cancer patients. Clin Cancer Res 2007;13: [16] Zhong W, Yang X, Bai J, Yang J, Manegold C, Wu Y. Complete mediastinal lymphadenectomy: the core component of the multidisciplinary therapy in resectable non-small cell lung cancer. Eur J Cardiothorac Surg 2008;34: APPENDIX. CONFERENCE DISCUSSION Scan to your mobile or go to to search for the presentation on the EACTS library Dr R. Schmid (Bern, Switzerland): You and your group from Nagasaki identified the problem of intraoperative staging when you do a sublobar resection. You are interested if you have to proceed to lobectomy, if you have a positive lymph node intraoperatively in the lobe. My first question is what is the time you need for the analysis? You do not have to wait for the pathologist, or as you said, you do not have a pathologist at hand in Japan at all. What is the timing of a complete intraoperative staging? Furthermore, which lymph nodes are you identifying or are you analysing? Are you only analysing interlobar lymph nodes, or also mediastinal lymph nodes at the same time? Dr Tomoshige: For the first question, the time it takes is about 40 minutes, so it is comparable to an intraoperative pathological diagnosis. Dr Schmid: Yes. Did you plan a study to compare pathological staging with your method intraoperatively? Because this would basically be the gold standard, to compare pathology results with your result. Dr Tomoshige: Yes. Before delivering to the pathological division, I cut the lymph node in two. Dr Schmid: Yes. So the pathologist also did intraoperative staging? Dr Tomoshige: Yes, of course. Dr Schmid: Not a postoperative analysis of the lobe? Dr Tomoshige: We performed these at the same time. Dr Schmid: Did you draw any conclusions from your intraoperative staging in this series? Did you change your operative strategy when you found positive lymph nodes?

7 K. Tomoshige et al. / European Journal of Cardio-Thoracic Surgery 7 Dr Tomoshige: No. We performed the operation with the intraoperative pathological diagnosis. Dr Schmid: You followed the pathological diagnosis? Dr Tomoshige: Yes. Dr Schmid: Is this technique also applicable with a thoracoscopic resection? Can you do a thoracoscopic lymph node sampling for this technique? Because many people go on to a thoracoscopic segmentectomy, or if necessary, a lobectomy, or did you do open surgery? Dr Tomoshige: In many cases, VATS, thoracoscopic surgery. Dr Schmid: So the intraoperative lymph node dissection was done by VATS? What was the percentage of open procedure and what was the percentage of VATS procedures you performed? Dr Tomoshige: It is a good option for VATS thoracoscopic surgery, yes. Dr P. Licht (Odense, Denmark): Could you just briefly comment on the sensitivity of your new method in detecting lymph node involvement? We do not know if this applies to any clinical outcome such as survival yet, but hopefully in the future, you will provide us some data. Would you think that your method is so sensitive, it can detect micrometastasis? And therefore, probably overestimate the nodal stage compared with conventional histology? Dr Tomoshige: This method has a high sensitivity, so we can detect micrometastasis, yes, of course. The concentration of color, if positive, depends on the size of the metastasis. The larger the size was, the more intense the colour became. Dr G. Kocher (Bern, Switzerland): Have you also tried or assessed your method in EBUS? Dr Tomoshige: Not yet. Dr Kocher: We would expect a high likelihood of false positive results, because if you puncture through the bronchial wall, you might even get some alveolar tissue in it. So you have not assessed if it is useful or not for EBUS? Dr Tomoshige: We did not.