ATS
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to Personal Folders
Right arrow Download to citation manager
Right arrow Author home page(s):
George Ladas
Peter Goldstraw
Right arrow Permission Requests
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Watanabe, S.-i.
Right arrow Articles by Goldstraw, P.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Watanabe, S.-i.
Right arrow Articles by Goldstraw, P.
Related Collections
Right arrow Lung - cancer
Right arrowRelated Article

Ann Thorac Surg 2002;73:245-248
© 2002 The Society of Thoracic Surgeons

Original article: general thoracic

Inter-observer variability in systematic nodal dissection: comparison of European and Japanese nodal designation

Shun-ichi Watanabe, MDa, George Ladas, FRCSa, Peter Goldstraw, FRCS*a

a Department of Thoracic Surgery, Royal Brompton Hospital, London, England, UK

Accepted for publication July 30, 2001.

* Address reprint requests to Dr Watanabe, Department of Surgery (I), Kanazawa University School of Medicine, 13-1 Takara-machi, Kanazawa 920-8641, Japan
e-mail: shunuk{at}aol.com


   Abstract
Top
 Abstract
Introduction
 Patients and methods
 Results
 Comment
 References
 
Background. Systematic nodal dissection is accepted as an important component of the intrathoracic staging of patients undergoing thoracotomy for lung cancer. Several lymph node maps have been proposed in an attempt to ensure uniformity in designating lymph node stations. The Japan Lung Cancer Society has published detailed definitions for each nodal station adopting the Naruke map. However, since these definitions had not been interpreted into other languages, they have not been universally accepted. The objective of this study was to assess the inter-observer variability in the interpretation of lymph node stations.

Methods. A total of 424 lymph node stations were removed from 41 patients undergoing thoracotomy for non-small cell lung cancer. All nodal stations were labeled using the Naruke map. As each station was excised, it was designated in a blind fashion by one of two surgeons trained in the UK and one surgeon trained in Japan. The designation accorded to each nodal station was analyzed.

Results. The total concordance was 68.5% (right side 67.0%, left side 69.9%). The concordance rate for individual nodal stations varied from 0% to 100%. Considerable discordance existed between the Japanese and European surgeons in the designation of nodal stations 2, 4, 8 and N1 station 12. In 14 (34.1%) patients, discordance in the labeling of lymph nodes led to disease being categorized as N1 by one observer, whereas the other considered the same nodes to be N2.

Conclusions. Considerable discordance in the designation of nodal station has been demonstrated. We would expect similar inter-observer variability elsewhere between surgeons, institutions, or countries. More detailed nodal charts and precise, easily understood definitions of nodal stations are needed for intrathoracic staging. The first English version of the Japan Lung Cancer Society staging manual goes some way to address this.


   Introduction
Top
 Abstract
 Introduction
Patients and methods
 Results
 Comment
 References
 
In 1978, Naruke and associates [1] suggested the use of an anatomical map in which the lymph node stations were numbered. The numbers 1 to 9 were used to designate mediastinal (N2) stations, and numbers 10 to 14 for N1 stations. This map was intended to be a visual aid to ensure that the various lymph node stations involved by tumor would be uniformly recorded in lung cancer patients. The Japan Lung Cancer Society adopted this map and established detailed definitions for each nodal station in their manual for the Classification of Lung Cancer. Although the map has been widely used outside Japan, the explanatory manual only became available in English in March 2000 [2]. There was obviously scope for surgeons to designate nodal stations differently when relying upon their visual interpretation of a chart without the detailed definitions contained in the accompanying manual. The objective of this study was to assess the extent and impact of any such inter-observer error. We hope this will facilitate the creation of a common map with internationally recognized definitions.


   Patients and methods
Top
 Abstract
 Introduction
 Patients and methods
Results
 Comment
 References
 
The subjects for this study were 41 consecutive patients undergoing thoracotomy for non-small cell lung cancer (NSCLC) at the Royal Brompton Hospital. Preoperative staging included computed tomography (CT) of the chest and abdomen in all patients. Cervical mediastinoscopy was performed in 15 patients (36.6%) in whom the CT scan had raised the possibility of mediastinal nodal involvement or mediastinal invasion. For tumors arising in the left upper lobe, or those which had extended to the left main bronchus, this evaluation was supplemented by left anterior mediastinotomy. Any patient confirmed to have N2 disease was excluded from the series. Preoperative staging for all patients was therefore cT1–3 N0–1 M0. At thoracotomy, the diagnosis was confirmed by frozen-section analysis if histological confirmation was not available preoperatively. Systematic nodal dissection was then performed prior to making any decision as to resectability. As a first step, all ipsilateral mediastinal nodal stations were removed and stored separately upon a sterile card. The macroscopic appearances and internal architecture of the nodes was assessed by serially sectioning each station. If necessary, frozen section examination of key nodes was performed. This evaluation was then repeated for the N1 nodes, extending peripherally in a centrifugal fashion until the surgeon considered that sufficient information was available to allow a decision as to the desirability of resection and the extent required. This allowed the surgeon to assess the feasibility and advisability of complete clearance before commencing resection.

For this study, two observers scrubbed for each case. One observer had been trained in Japan and was familiar with the Naruke chart and the Classification of Lung Cancer issued by the Japan Lung Cancer Society. The other observer was one of two surgeons trained in the UK: One had been using the Naruke chart for 20 years without formal training, and the other had trained with that surgeon for 5 years. This study predated the first English edition of the staging manual [2]. As each nodal station was excised, it was labeled alphabetically. Both observers then made their own interpretation of the nodal designation, writing this onto a sheet of paper that could not be seen by the other observer. If frozen section was requested, the alphabetical designation was used. The work sheets from the two observers were filed in a sealed envelope for each case. These were opened once the study had been completed.

Statistical analysis
All values were expressed as means ± SD (standard deviation). The statistical significance of differences was determined using unpaired t-test, {chi}2 test for independence, and one-factor analysis of variance. Relative risk and 95% confidence intervals were calculated. Values of p less than 0.05 were considered to be statistically significant.


   Results
Top
 Abstract
 Introduction
 Patients and methods
 Results
Comment
 References
 
Patient characteristics
The data was complete for all 41 cases. Patient characteristics were shown in Table 1. Right thoracotomy was performed in 20 patients and left thoracotomy in 21. The lobe of origin was the right upper lobe (RUL) in 10 patients, right middle lobe (RML) in 2, right lower lobe (RLL) in 8, left upper lobe (LUL) in 14, and left lower lobe (LLL) in 7. Open and close thoracotomy was performed in 2 patients (5%). Resection was undertaken in the other 39 patients, by pneumonectomy in 7, bilobectomy in 1, lobectomy in 28 (of whom 21 underwent conventional lobectomy, 5 bronchial sleeve-resection, and 2 broncho-vascular double-sleeve lobectomy), segmentectomy in 2, and wedge resection in 1 (Table 1).


View this table:
[in this window]
[in a new window]
  		Table 1. Patient Characteristics

 
Removed lymph nodes
Four hundred and twenty-four lymph node stations were removed in these 41 patients, an average of 10.4 ± 2.3 stations per person (range 5 to 15) (Table 1). For 20 patients undergoing right thoracotomy, the average number of nodal stations examined was 10.1 ± 2.3, and for 21 undergoing left thoracotomy this number was 10.7 ± 2.2, with no significant difference (p = 0.388). The average number of stations removed per person was 10.5 ± 1.2 in RUL, 7.5 ± 0.7 in RML, 10.1 ± 3.2 in RLL, 10.6 ± 2.3 in LUL, and 11.0 ± 2.5 in LLL. The number was not influenced by the lobe of origin (p = 0.459) (Table 1). The average number of stations removed per person grouped by surgical procedure was 9.3 ± 3.0 in pneumonectomy, 10.0 in bilobectomy, 10.7 ± 2.3 in conventional lobectomy, 10.3 ± 1.1 in sleeve lobectomy, 12.5 ± 2.1 in segmentectomy, 12.0 in wedge resection, and 8.5 ± 0.7 in exploratory thoracotomy. The number was not influenced by the type of surgical procedure (p = 0.479) (Table 1). However, as one would expect, there was a tendency for more nodal stations to have been removed before any decision to undertake a more limited resection.

Eight nodal stations were excluded from analysis as they were designated as local variants of the Naruke chart for which no definition existed in the Japan Lung Cancer Manual. Of the remaining 416 nodal stations, 177 were designated N1 by the UK surgeon, 229 as N2, and 10 as N3. Those nodes designated as N3 were not considered for further analysis due to the small numbers involved.

Concordance rate
The overall concordance rate, ie, the number of stations in which there was agreement as a percentage of the total number of dissected stations, was 68.5% (285 out of 416). The concordance rate for right thoracotomy was 67.0% (132 out of 197), and for left thoracotomy was 69.9% (153 out of 219). There were no significant differences between these two groups (p = 0.212). The concordance rate for tumors of RUL origin was 57.8%, RML 73.2%, RLL 68.7%, LUL 74.5%, and LLL 64.9%. The concordance was not influenced by the lobe of origin (p = 0.325).

Table 2 shows the concordance rate for N1 stations. The overall concordance rate for N1 stations was 72.3% (128 of 177). There was clearly disagreement in the designation of all of the N1 stations, but this was particularly marked for station #12. The overall concordance rate for this station was only 60% (24 of 40), the Japanese surgeon considering that in 17.5% (7 of 40) of cases, the correct designation was #11, while in another 15.0% (6 of 40), it should have been #10.


View this table:
[in this window]
[in a new window]
  		Table 2. Concordance for N1 Stations

 
Table 3 shows the overall concordance rate for mediastinal nodal stations and for individual N2 stations. There was concordance in 65.1% of N2 stations (149 of 229). There was disagreement involving nearly all N2 nodal stations, with concordance rates varying from 0% to 100%. The disagreement was particularly noticeable for stations #2, 4, and 8. The Japanese surgeon considered 73.7% (14 of 19) of nodes labeled as #2 by the UK surgeons should have been designated as station #3, 22.5% (9 of 40) designated as station #4 should have been #3, and 39.5% (15 of 38) of station #8 should have been #9.


View this table:
[in this window]
[in a new window]
  		Table 3. Concordance for N2 Stations

 
The overall concordance rates between each UK surgeon and the Japanese surgeon were 68.0% (202 of 297) and 69.7% (83 of 119), with no significant difference (p = 0.731).

Discordance between N1 and N2 stations
More worrying was the discordance in which one surgeon considered a nodal station to be mediastinal in origin, while the other observer thought it to be of N1 origin. This occurred in 14 patients (34.1%), 6 having right thoracotomy and 8 left thoracotomy (Table 4). In 10 of these cases, the disagreement centered upon the distinction between N1 station #10 and N2 station #4, but there was also confusion between station #10 and station #5 in four left-sided cases.


View this table:
[in this window]
[in a new window]
  		Table 4. Discordance Between N1 and N2 Stations

 

   Comment
Top
 Abstract
 Introduction
 Patients and methods
 Results
 Comment
References
 
Surgeons have long been aware that the situation at thoracotomy is not always as was predicted by preoperative investigations. In recent decades, the intrathoracic reevaluation of disease at thoracotomy for lung cancer has evolved into a detailed and sophisticated assessment of disease extent. Central to this is an evaluation of nodal involvement at mediastinal and hilar levels. This technique, now termed "systematic nodal dissection" (SND), has been accepted by the International Association for the Study of Lung Cancer (IASLC) to be an important component of intrathoracic staging [3–5]. Several studies have shown that the positive and negative predictive values for CT in assessing mediastinal nodal involvement is in the order of 60% to 65% [6–8]. One of our earlier studies [4] suggested that SND could disclose "unexpected" N2 disease irrespective of cell-type, size, location, and lobe of origin of the primary tumor, and whether or not prior mediastinoscopy had been performed. We concluded that SND remained an important investigative process in all patients coming to surgery for lung cancer. Others have gone further, and suggested that cure rates may be improved by SND [9–11].

Systematic nodal dissection involves the identification of nodal stations and their labeling in accordance with an internationally recognized nodal chart. Several lymph node maps have been proposed [1, 12], each with their advantages and disadvantages. The one most widely used is that proposed by Naruke in 1978 [1]. In 1978, the Japan Lung Cancer Society published the first detailed definitions of each nodal station. This manual provided a meticulous definition for each station, based on CT and surgical findings, and was intended for clinical use. Although widely used within Japan [13], it has not become accepted elsewhere. Undoubtedly the major reasons for this was the lack of an English translation until March 2000 [2], but it has to be said that some of the definitions are less than totally clear.

Our study showed the concordance rate for individual nodal stations varied enormously, from 0% to 100%. Considerable discordance existed in the designation of stations #2, 4, 8, and N1 stations #12. As shown in Table 2, station #2 had very low (5.4%) concordance rate. More than 90% of station #2 labeled by UK surgeons were designated #3 by a Japanese surgeon. According to the Japanese manual, the boundary between station #2 and #3 is the pretracheal line. This would be appropriate in preoperative staging in CT films. However, in the surgical setting, there could occasionally be some difficulties in designating nodal stations, especially in the superior mediastinum. If the patient had undergone mediastinoscopy prior to thoracotomy, the procedure itself and resultant hematoma could displace upper mediastinal stations. In our series, 36.6% of patients had undergone mediastinoscopy to exclude N2 disease a few days before thoracotomy. This might have led to the confusion between pre- (#3) and paratracheal (#2) stations based on the Japanese manual in this group of patients. The American Joint Committee on Cancer nodal chart avoids such controversy simply because it does not discriminate between stations anterior to the trachea and those lying lateral to it.

Our study showed other disagreements on the precise position of nodes at "watershed" areas such as #1 versus #2 or #3, #4 versus #10, #8 versus #9, and further out into hilum, #12 versus #13. In each case, the Japanese manual relates the center of the node with some straight lines drawn on the anatomical structures to determine the designation. However, to define mediastinal lymph nodes accurately sometimes seems to be very difficult especially in the case of lymphadenopathy or irregular-shaped nodes. This confusion would be exacerbated by edema or hematoma, and where there are anatomical variations in the anatomical landmarks.

While any discordance in the designation of nodal stations will impact upon our perception of nodal spread, the differences become even more important when they concern the distinction between N2 and N1 stations. Discordance in this situation could, if these nodes were involved, distort the reported stage of patients in differing studies. Predictably, in our study such variation involved the "interface" nodes at the junction of hilum and mediastinum, the distinction between #10 and #4. Although the Japanese rulebook defines station #4 as just "medial to the azygos vein" this still leaves some confusion as to the upper and lower boundaries of this station. Actually, our study showed 34.1% of discordance between N1 and N2 stations as shown in Table 4. If we are to resolve the long-standing arguments relating to stations #4 and #10, we need to have clarity when depicting anatomy on nodal charts.

We found no differences in nodal designation between the two UK surgeons. The concordance rate between each UK surgeon and the Japanese surgeon being 68.0% and 69.7%. One of the UK surgeons was trained by the other, suggesting that the variability we found with nodal designation was not inherent within the system but reflected differences in interpretation which could be taught, and therefore perpetuated within different training centers.

There is a clear need for improved nodal charts, providing more detail in an unambiguous way. Newer educational techniques should be explored, perhaps combining computer-assisted design and three-dimensional holography.


   References
Top
 Abstract
 Introduction
 Patients and methods
 Results
 Comment
 References
 

  1. Naruke T., Suemasu K., Ishikawa S. Lymph node mapping and curability at various levels of metastasis in resected lung cancer. J Thorac Cardiovasc Surg 1978;76:832-839.[Abstract]
  2. Japan Lung Cancer Society. Classification of lung cancer. First English edition. Tokyo: Kanehara & Co, Ltd, 2000.
  3. Naruke T., Tsuchiya R., Kondo H., Asamura H., Nakayama H. Implications of staging in lung cancer. Chest 1997;112(Suppl 4):2421-2428.
  4. Graham A.N., Chan K.J., Pastorino U., Goldstraw P. Systematic nodal dissection in the intrathoracic staging of patients with non-small cell lung cancer. J Thorac Cardiovasc Surg 1999;117:246-251.[Abstract/Free Full Text]
  5. Ginsberg R.J. Continuing controversies in staging NSCLC: an analysis of the revised 1997 staging system. Oncology 1998;12(Suppl 2):51-54.[Medline]
  6. Glazer G.M., Gross B.H., Quint L.E., Francis I.R., Bookstein F.L., Orringer M.B. Normal mediastinal lymph nodes: number and size according to American Thoracic Society mapping. Am J Roentgenol 1985;144:261-265.[Abstract/Free Full Text]
  7. McLoud T.C., Bourgouin P.M., Greenberg R.W., et al. Bronchogenic carcinoma: analysis of staging in the mediastinum with CT by correlative lymph node mapping and sampling. Radiology 1992;182:319-323.[Abstract/Free Full Text]
  8. Glazer G.M., Orringer M.B., Gross B.H., Quint L.E. The mediastinum in non-small cell lung cancer: CT-surgical correlation. Am J Roentgenol 1984;142:1101-1105.[Abstract/Free Full Text]
  9. Watanabe Y., Shimizu J., Oda M., et al. Improved survival in left non-small cell lung cancer after more extensive operative procedure. Thorac Cardiovasc Surg 1991;39:89-94.[Medline]
  10. Pearson F.G. Non-small cell lung cancer. Role of surgery for stage I–III. Chest 1999;116(Suppl 6):500-503.
  11. Keller S.M., Adak S., Wagner H., Johnson D.H. Mediastinal lymph node dissection improves survival in patients with stage II and IIIa non-small cell lung cancer. Ann Thorac Surg 2000;70:358-366.[Abstract/Free Full Text]
  12. Mountain C.F., Dressler C.M. Regional lymph node classification for lung cancer staging. Chest 1997;111:1718-1723.[Abstract/Free Full Text]
  13. Naruke T., Tsuchiya R., Kondo H., Nakayama H., Asamura H. Lymph node sampling in lung cancer: how should it be done?. Eur J Cardiothorac Surg 1999;16(Suppl 1):17-24.

Related Article

Invited commentary
Tsuguo Naruke
Ann. Thorac. Surg. 2002 73: 248-249. [Extract] [Full Text] [PDF]



This article has been cited by other articles:


Home page
 Ann. Thorac. Surg.Home page
Y. Shimada, M. Tsuboi, H. Saji, K. Miyajima, J. Usuda, O. Uchida, N. Kajiwara, T. Ohira, T. Hirano, H. Kato, et al.
The prognostic impact of main bronchial lymph node involvement in non-small cell lung carcinoma: suggestions for a modification of the staging system.
Ann. Thorac. Surg., November 1, 2009; 88(5): 1583 - 1588.
[Abstract] [Full Text] [PDF]

Home page
 Eur. J. Cardiothorac. Surg.Home page
W. Zhong, X. Yang, J. Bai, J. Yang, C. Manegold, and Y. Wu
Complete mediastinal lymphadenectomy: the core component of the multidisciplinary therapy in resectable non-small cell lung cancer.
Eur. J. Cardiothorac. Surg., July 1, 2008; 34(1): 187 - 195.
[Abstract] [Full Text] [PDF]

Home page
 Eur. J. Cardiothorac. Surg.Home page
E. Belcher and P. Goldstraw
Is radical mediastinal dissection mandatory for curative resection of NSCLC?
Eur. J. Cardiothorac. Surg., January 1, 2007; 31(1): 142 - 143.
[Full Text] [PDF]

Home page
 Ann. Thorac. Surg.Home page
D. Lardinois, H. Suter, H. Hakki, V. Rousson, D. Betticher, and H.-B. Ris
Morbidity, Survival, and Site of Recurrence After Mediastinal Lymph-Node Dissection Versus Systematic Sampling After Complete Resection for Non-Small Cell Lung Cancer
Ann. Thorac. Surg., July 1, 2005; 80(1): 268 - 275.
[Abstract] [Full Text] [PDF]

Home page
 J. Thorac. Cardiovasc. Surg.Home page
M. Okada, T. Sakamoto, T. Yuki, T. Mimura, H. Nitanda, K. Miyoshi, and N. Tsubota
Border between N1 and N2 stations in lung carcinoma: Lessons from lymph node metastatic patterns of lower lobe tumors
J. Thorac. Cardiovasc. Surg., April 1, 2005; 129(4): 825 - 830.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to Personal Folders
Right arrow Download to citation manager
Right arrow Author home page(s):
George Ladas
Peter Goldstraw
Right arrow Permission Requests
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Watanabe, S.-i.
Right arrow Articles by Goldstraw, P.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Watanabe, S.-i.
Right arrow Articles by Goldstraw, P.
Related Collections
Right arrow Lung - cancer
Right arrowRelated Article

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
ANN THORAC SURG ASIAN CARDIOVASC THORAC ANN EUR J CARDIOTHORAC SURG
J THORAC CARDIOVASC SURG ICVTS ALL CTSNet JOURNALS