ARTICLE IN PRESS

www.icvts.org

doi:10.1510/icvts.2010.236349

Interactive CardioVascular and Thoracic Surgery 11 (2010) 260–264

� 2010 Published by European Association for Cardio-Thoracic Surgery

State-of-the-art - Thoracic oncologic

Prognostic classifications of lymph node involvement in lung cancerand current International Association for the Study of Lung Cancer

descriptive classification in zonesMarc Riquet*, Alex Arame, Christophe Foucault, Francoise Le Pimpec Barthes

Department of Thoracic Surgery, Georges Pompidou European Hospital, 20 rue Leblanc, 75015 Paris, France

Received 24 February 2010; received in revised form 27 May 2010; accepted 2 June 2010

Summary

The lymphatic drainage of solid organ tumors crosses through the lymph nodes (LNs) whose tumoral involvement may still be consideredas local disease. Concerning lung cancer, LN involvement may be intrapulmonary (N1), and mediastinal andyor extra-thoracic. More than30 years ago, mediastinal involved LNs were all considered as N2, and outside the scope of surgery. In 1978, Naruke presented an originalarticle entitled ‘Lymph node mapping and curability at various levels of metastasis in resected lung cancer’, demonstrating that N2 wasnot a contraindication to surgery in all patients. The map permitted to localize the favorable N2 on the lung cancer ipsilateral side of themediastinum. Several maps ensued aiming to discriminate between right and left involvement (1983), and to distinguish N2 (ipsilateral)and N3 (contralateral) mediastinal LN involvement (1983, 1986). The last map (1997 regional LN classification) was recently replaced by adescriptive classification in anatomical zones. This new LN map of the TNM classification for lung cancer is a step toward using anatomicalview points which might be the best way to better understand lung cancer lymphatic spread. Nowadays, the LNs are easily identified bycurrent radiological imaging, and their resectability may be anticipated. Each LN chain may be removed by en-bloc lymphadenectomyperformed during radical lung resection, a safe procedure which seems to be more oncological based than sampling, and which avoids thesource of discrepancies pointed out during the labeling of LN stations by surgeons.� 2010 Published by European Association for Cardio-Thoracic Surgery. All rights reserved.

Keywords: Lung cancer; TNM; N1; N2; Surgery; Staging; Lymphatic drainage

The lymphatic drainage of solid organ tumors crossesthrough the lymph nodes (LNs), whose tumoral involvementmay still be considered as local disease. The TNM classifi-cation stresses this important oncological point by individ-ualizing the N, so as to indicate whether the LNs areinvolved or disease free. In effect, removing the tumor andthe involved LN may offer a chance of a cure, whereas Mpositive status already represents a systemic disease. LNinvolvement in lung cancer, is always an indicator of apoorer prognosis. The LNs may be unresectable as a resultof invasion to the adjacent structures or their wide disper-sion. Indeed, the more the degree of involvement, themore are the chances for the tumor to be associated withdistant systemic metastases, rendering resection useless.The challenge is to find relevant criteria to estimatewhether surgical resection is worthwhile or not, andyorwhether neoadjuvant therapy may render the tumor ame-nable to surgery in the hope of a cure.

With regards to lung cancer, LN involvement may beintrapulmonary (N1), easily resectable with the lung as awhole, mediastinal (N2), and extra-thoracic or mediastinalyhilar contralateral (N3), thus contra-indicating surgery.

*Corresponding author. Tel.: q33 156 093450; fax: q33 156 093380.E-mail address: marc.riquet@egp.aphp.fr (M. Riquet).

More than 30 years ago, lung cancer included carcinoidtumors, small cell cancers and non-small cell lung cancers.Mediastinal either ipsi- or contralateral involved LNs wereall considered as N2, and outside the scope of surgery.

In 1978, Naruke et al. w1x presented an original articleentitled ‘Lymph node mapping and curability at variouslevels of metastasis in resected lung cancer’. Out of 468patients who underwent surgery, 270 underwent curativeor near curative operations, among 64 of whom had N2involvement. The location of these LNs was reportedaccording to a map of 14 LN stations. The removed N2proved to be located in the tumor entirely in the ipsilateralpart of the mediastinum. Fifteen patients remained alive,indicating a 18.8% five-year survival rate. It was concludedthat N2 was not a contraindication to surgery in all patients.In 1980, Martini et al. w2x reported 445 ‘N2’ lung cancers.A complete, potentially curative resection was possible for80 patients. LNs location was reported using a map of 13LN stations available in the 1979 brochure of the AmericanJoint Committee (AJC), whose concept was similar toNaruke results. Their results on survival confirmed Naruke’sone.

Identifying the subset of N2 patients likely to benefit fromsurgery thus appeared unequivocal, but the AmericanThoracic Society (ATS) questioned the anatomical validity

Downloaded from https://academic.oup.com/icvts/article-abstract/11/3/260/813245by gueston 18 March 2018

ARTICLE IN PRESS

261M. Riquet et al. / Interactive CardioVascular and Thoracic Surgery 11 (2010) 260–264

New

IdeasInstitutional

ReportW

orkin

ProgressReport

ESCVSArticle

NegativeResults

State-of-the-artBest

EvidenceTopic

BriefCom

munication

CaseReport

Follow-up

PaperEditorial

ProtocolProposalfor

Bail-out

ProcedureN

omenclature

HistoricalPages

of the mapping w3x. Several of the LN stations of Narukewere judged to be defined in non-specific terms and opento subjective interpretation, and the various LNs in the AJCmap appeared as labeled dots on a diagram, making thelimits of each station non-specific and highly subjectivew3x. Therefore, the above-mentioned survival data, theadvances observed in some diagnostic techniques (medias-tinoscopy and tomodensitometry), and the lack of defini-tion of each LN station, prompted the ATS to propose amodified clinical staging of primary lung cancer in 1983.The new presentation of the AJC mapping took into accountthe major anatomical structures that could be identifiedat mediastinoscopy, and distinguished the right from theleft mediastinal LN involvement, considering subcarinal LNinvolvement as bilateral involvement. In 1986, Mountainproposed a ‘new international staging system for lungcancer’ w4x, that maintained the 1983 mapping but termedN2 the ipsilateral mediastinal LN involvement, as betteroutcome was anticipated for such patients, and N3 thesupraclavicular, opposite hilum, and contralateral LNinvolvement whose prognosis was extremely poor. However,N3 designation was assigned for all these LN stationsbecause they were usually included in the regional field ofradiation therapy, and not because of anatomicalconsiderations.

At that time, Martini et al. w5x and Libschitz et al. w6x,reported the existence of skipping NSCLC metastases, thatis mediastinal involved LNs, without intrapulmonary dis-eased ones. A study demonstrating the existence of lunglymphatic vessels running directly to the mediastinal nodeswithout encountering intrapulmonary LN on their way tothe mediastinum w7, 8x provided an anatomical explanationfor that skipping mode of spread. The study also describedthe LN chains draining the lungs and involved by non-smallcell lung cancer (NSCLC) w8x. The latter study, performedon a greater series of adult cadavers, was compared withthe N2 pattern of patients who underwent surgery w9x. Theanatomical description of the LN chains proved to be mainlyequivalent to Rouviere’s 1929 description w10x, but outlinedthe 1983 AJC-map lack of precision. The correlated clinicalstudy w9x demonstrated that prognosis was better whenonly one mediastinal LN chain was involved (five-yearsurvival rates 27.4% vs. 9.31% when two or more chains areinvolved, P-0.0001). An explanation was suggested by theanatomical study: the lymph flowing within the chains wasdemonstrated to drain towards the venous circulation inthe neck, and also to enter the thoracic duct in themediastinum, and to connect with the ipsi- and contralat-eral LN chains. Single involved LN chain was still a localdisease in the 27% of patients who survived, whereas theothers were already suffering from non-detectable systemicmetastases. For the patients with two or more involvedchains, the risk of metastazing tumor cells in the circulationand future implantation was dramatically increased by toomany connections with the blood circulation, and explainsthe poorer prognosis. Systemic metastases as the majorcause of cancer related deaths (90% of patients, 117y130)supported this finding.

However, there was no data to explain why N1 survivalrates were always between those observed for N0 and N2patients in the literature. Five-year survival rate was

reported to be significantly better in the case of intralobarN1 than in the case of extralobar-hilar N1 w11, 12x. Intra-lobar N1 five-year survival rate was demonstrated to besimilar to that of N0, and extralobar N1 to five-year survivalrate of N2 involving a single LN chain w13x. The explanationhas anatomical basis: extralobar-hilar (stations 10 and 11)LNs are the first nodes of the LN chain w10x, whereasintralobar LNs may appear after birth w14x, probably origi-nating from the bronchial-associated lymphoid tissue(BALT) w15x.

The 1997 ‘regional LN classification for lung cancer stag-ing’ provided the same map as in 1983, but suppressed thelines indicating the anatomical borders of the stations w16x.Soon after, Asamura et al. w17x, questioned the boundarybetween N1 and N2 stations and suggested incorporatingthe LNs around the main bronchus in N2. They reportedfindings concerning N1 similar to those previously men-tioned w11–13x. This confirmed the validity of resorting toanatomy for pathological N classification w18x: N0 includingintralobar N1, N1 including extralobar-hilar LN and singleinvolved mediastinal LN chain, and N2 being two or moremediastinal LN chain involvement w18x. Similar results con-firmed this prognostic significance of main bronchial LNinvolvement (station 10) w19, 20x, and other authors con-firmed that single station N2 involvement has a betterprognosis than multiple station N2 w21, 22x. In the case ofsingle station N2 involvement, only a single mediastinal LNchain could be involved.

In 2007, the International Association for the Study ofLung Cancer (IASLC) published proposals for the revision ofthe N descriptors in the forthcoming seventh edition of theTNM classification for lung cancer w23x. They were theresults of immense work including 47 databases, 38,265patients with no clinical evidence of metastatic disease ofwhom 28,371 were managed surgically. However, only 12out of the 47 databases included data on sampling results(positive, negative, not done) for the individual nodalstations, and pathologic N1 (pN1) and pathologic N2 (pN2)stage were analyzed for 2876 patients. On the basis of theavailable data, no proposal altering the current N stagedescriptors, was recommended. During the same year, areview of the lymphatics of the lungs and anatomicalmediastinal LN chains mainly based on the previous anatom-ical studies was provided w24x (Figs. 1 and 2).

In the right upper mediastinum, there are three LN chains.The right pre- or paratracheal LNs chain w5x has abundant

LNs, whose size decreases from below to upward and whosenumber varies from one individual to another. It corres-ponds to the stations 4R, 2R, and superior 10R w3x or LN ofthe azygos vein arch of Rouviere w10x. The chain lies in thepretracheal space. The pretracheal space is bounded anter-olaterally by the superior vena cava, right brachio-cephalic vein and mediastinal pleura on the right, the aortaand pericardium on the left, and posteriorly by the tracheaw25x.

The tracheo-esophageal chain w5x consists of a lymphvessel with nodes at the thoracic inlet of the mediastinum;rarely, a retrobronchial node is present. It corresponds tostation 3p w3x. The chain lies in the right tracheo-esopha-geal groove.

Downloaded from https://academic.oup.com/icvts/article-abstract/11/3/260/813245by gueston 18 March 2018

ARTICLE IN PRESS

262 M. Riquet et al. / Interactive CardioVascular and Thoracic Surgery 11 (2010) 260–264

Fig. 1. Showing the relation between each lymph node chain, lymph nodestations and the zones of current IASLC mapping: supraclavicular zone: 1Rand 1L (two chains); upper zone: 2R and 4R (one chain); 2L and 4L (onechain); 3a (one chain); AP zone: 5 and 6 (one chain); subcarinal zone: 7 and8 (one chain); lower zone: 9 (one chain); hilaryinterlobar zone: 10 and 11.IASLC, International Association for the Study of Lung Cancer; AP, aorto-pulmonary.

Fig. 2. Lateral view showing the 3p and 3a lymph node stations within theirchain (upper zone of the IASLC mapping: 3p and 3a, one chain each). IASLC,International Association for the Study of Lung Cancer.

The right phrenic nerve chain runs along the right phrenicnerve and is equivalent to station 3a. It rarely participatesin the lymphatic drainage of the lung w3, 10x.

In the left upper mediastinum, there are also three LNchains.

The pre-aorto-carotid chain w5x: begins in front of theaortopulmonary space with a node commonly located infront of the ligamentum arteriosum, and ascends subpleu-rally in the triangle formed by the hilum of the lunginferiorly, the phrenic nerve anteriorly, and the vagus nerveposteriorly, the whole being covered by the mediastinalpleura w25x. LNs are not numerous. It corresponds to sta-tions 5 and 6 w3, 15x.

The left superior bronchial-recurrent chain w5x usuallyconsists of one or two lymphatics ascending in the lefttracheo-esophageal groove. It rarely contains nodes afterthe left superior bronchial (10L) and lower tracheal (4L)ones. It corresponds to stations (superior) 10L, 4L, and 2Lw3, 15x. 4L and 10L are located in the aorto-pulmonaryspace, commonly called as aorto-pulmonary window. It isbounded superiorly by the aortic arch, medially by the

trachea and esophagus, inferiorly by the superior border ofthe main left bronchus, anteriorly and laterally by thepulmonary artery and ligamentum arteriosum w25x.

The left phrenic nerve chain runs along the left phrenicnerve and is equivalent to station 3a. It rarely participatesin the pulmonary lymphatic drainage, but does so moreoften than the right side as a result of its close connectionwith the pre-aorto-carotidian chain w3, 10x.

In the lower mediastinum, there are three chains.A common chain for both lungs of intertracheobronchial

LNs w5x: has a variable number of nodes whose size variesfrom one individual to another and are arranged in threeclusters: one median subcarinal cluster flanked by rightand left clusters of nodes situated under the correspondingbronchus. These clusters correspond to stations 7, 8, andinferior 10R and 10L w3, 15x. The chain is located withinthe subcarinal space w25x, which is bounded superiorly bythe carina, laterally by the mainstem bronchi, anteriorlyby the pulmonary artery and the posterior aspect of thepericardium as well as left atrium, and posteriorly by theesophagus.

The inferior pulmonary ligament has a LN chain in eachside. LNs are neither frequent nor numerous, but lymphaticvessels are always present at this level. These chains arenear the lateral aspect of the esophagus, which is surround-ed by juxtaesophageal LN which mainly receives the lymphfrom the esophagus w26, 27x. However, the juxtaesophagealLNs do not normally participate in the lymph drainage ofthe lungs w5x.

The IASLC, by analyzing the N descriptors in the interna-tional database w23x, highlighted important differences innomenclature between the Japaneses (Naruke) LN map,which is still used by the Japanese Lung Cancer Society,and the 1997 ATS map (Mountain and Dresler) used in NorthAmerica and Europe. These differences, regarding theirreconcilable discrepancy in the data analysis and in the

Downloaded from https://academic.oup.com/icvts/article-abstract/11/3/260/813245by gueston 18 March 2018

ARTICLE IN PRESS

263M. Riquet et al. / Interactive CardioVascular and Thoracic Surgery 11 (2010) 260–264

New

IdeasInstitutional

ReportW

orkin

ProgressReport

ESCVSArticle

NegativeResults

State-of-the-artBest

EvidenceTopic

BriefCom

munication

CaseReport

Follow-up

PaperEditorial

ProtocolProposalfor

Bail-out

ProcedureN

omenclature

HistoricalPages

labeling of LN stations among surgeons, led to maintain thecurrent N descriptors in the TNM staging system w23x. As aconsequence, a revised LN map was proposed, putting asidethe main differences between the Japanese and ATS maps,which aimed to provide more specific anatomical defini-tions for each of the LN stations. The main differencesincluded: station 1 in the Naruke map corresponding tostations 1 and 2 in the ATS map; stations 2, 3, 4R and 4Lin the Naruke map corresponding to stations 4R and 4L inthe ATS map; stations 10 and 7 in the Naruke map corre-sponding to station 7 in the ATS map. The new internationalLN map was published in 2009 w28x, and adopted in theseventh TNM classification w29x. It describes the new LNstations in several ‘anatomical zones’ (ns7) w28x. Groupingthe LN stations into ‘anatomical zones’ was the result ofstatistical exploratory analyses used to reconcile the Naru-ke and ATS LN maps, and not the result of direct comparisonwith human anatomy. The LN station description resemblesthe LN chain anatomy, but is still inaccurate. The remainingpoints concerning anatomy to make the prognostic value ofthe defined anatomical zones clear, need furtherinformation.

The discrepancies between stations 2 and 4 have not beencompletely resolved. In fact, 2R and 4R pertain to the pre-or paratracheal LN chain, and 2L and 4L to the left superiorbronchial-recurrent chain: there are four stations, but twochains from a strictly anatomical point of view. The arbi-trary division along the midline of the trachea created bythe ATS has been rightly eliminated: it was also not ana-tomical, and all LN in the pretracheal space pertain to thepre- or paratracheal chain, obviously including the Narukestation 3 overlying the midline of the trachea.

The entire subcarinal group of LN is now defined as awhole, which perfectly corresponds to anatomy. However,station 8 (paraesophageal) is also a part of this group, andmay not be, or may only occasionally be, correlated withstation 9.

Stations 5 and 6 pertain to the same chain (pre-aorto-carotid chain). The left bronchial–recurrent chain (4Lq2L)is an important lymphatic pathway from the left lung, andis as important as the pre-aorto-carotid chain: 2L nodesare rare but 4L nodes are very frequent, and are deeplylocated in the aortopulmonary space.

The right tracheo-esophageal chain (3p), and the rightphrenic nerve chain (3a) are rarely concerned by thelymphatic drainage of the lungs. Frequency of N2 involve-ment was 1.7% for each of both (2y116 N2 patients) w9x.

Grouping the LN stations into the anatomical zones sug-gested by exploratory analyses and proposing the latter asindicators of prognosis indirectly allowed the N0, N1, andN2 classification reappraisal to be deduced from anatomo-clinical correlations w9, 18x. Effectively, each zone mainlycorresponds to a single LN chain. The upper zone is thepre-paratracheal chain N2 for the right lung and bronchial-recurrent chain N2 for the left lung; 3a and 3p are toorarely involved to change prognosis when using exploratorystatistical analyses. However, 3a or 3p positivity must beconsidered as one different anatomical LN chain. Aorto-pulmonary (AP) zone is the pre-aorto-carotid chain. Thesubcarinal zone corresponds to the nodes of the bifurcation,and includes station 8. The lower zone is the inferior

pulmonary ligament LN chain (paraesophageal nodes beingrarely involved and usually it is in the subcarinal zone).The hilaryinterlobar zone is correctly defined, both LNstations having the prognosis value of a single LN chain.However, the role of the peripheral zone is the onlyremaining problem. It is difficult to distinguish intralobarstation 12 from interlobar station 11. In the maps publishedby Naruke et al. w1x, Mountain and Dresler w16x, and IASLCw28x, the circles with 12 and 11 are either at the samelevel, either slightly overlapping each other, but rarelypresented separately. In the same manner, it is difficult forthe surgeon who is resecting LNs located at that level, tostate whether the LNs are station 12 the LN having devel-oped towards the interlobar region, or station 11 the LNsdeveloped towards the lobar bronchus. The five-year sur-vival rate of stations 10 and 11 N1 was not statisticallydifferent from the rate of single station N2: however, thesurvival curves were not completely juxtaposed, and thestations 10 and 11 curve were slightly displaced towardsthe five-year survival curves of stations 13 and 14 N1 andN0 w13x. This is probably due to including station 12 LNs inthe station 11.

At the beginning of lung cancer story, mediastinal LNinvolvement was of major concern, precluding surgery, anddifficult to understand. Over time experience showedpatients whose LN involvement could be cured by surgery.This LN involvement was classified according to practicalexperience and represented mappings evolving over timeand moving step by step closer to anatomical landmarks.The current LN map of the TNM classification for lungcancer is a further step returning a little more to anatomy.It is worth stressing that entirely resorting to appliedanatomy should probably be the best way for better under-standing of lymphatic mode of spread of lung cancer.Nowadays, the LNs are easily identified by current radiologyimaging, and their resectability may be anticipated. EachLN chain may be removed by en-bloc lymphadenectomyperformed during radical lung resection, a safe procedurewhich is more oncological than sampling, which avoids thesource of misleading discrepancies pointed out during thelabeling of LN stations by surgeons.

References

w1x Naruke T, Suemasu K, Ishikawa S. Lymph node mapping and curabilityat various levels of metastasis in resected lung cancer. J ThoracCardiovasc Surg 1978;76:833–839.

w2x Martini N, Flehinger BJ, Zaman MB, Beattie EJ. Prospective study of445 lung carcinomas with mediastinal lymph node metastases. J ThoracCardiovasc Surg 1980;80:390–399.

w3x American Thoracic Society (Tisi GM, Friedman PJ, Peters RM, PearsonG, Care D, Lee RE, Selawry O). Clinical staging of primary lung cancer.Am Rev Respir Dis 1987;127:659–669.

w4x Mountain CF. A new international staging system for lung cancer. Chest1986;89:225S–233S.

w5x Martini N, Flehinger BJ, Zaman MB, Beattle EJ Jr. Results of resectionin non-oat cell carcinoma of the lung with mediastinal lymph nodemetastases. Ann Surg 1983;198:386–397.

w6x Libschitz HI, McKenna RJ, Mountain CF. Patterns of mediastinal metas-tases in bronchogenic carcinoma. Chest 1986;90:229–232.

w7x Riquet M, Hidden G, Debesse B. Direct lymphatic drainage of lungsegments to the mediastinal nodes. An anatomic study on 260 adults.J Thorac Cardiovasc Surg 1989;97:623–632.

Downloaded from https://academic.oup.com/icvts/article-abstract/11/3/260/813245by gueston 18 March 2018

ARTICLE IN PRESS

264 M. Riquet et al. / Interactive CardioVascular and Thoracic Surgery 11 (2010) 260–264

w8x Riquet M. Anatomic basis of lymphatic spread from carcinoma of thelung to the mediastinum: surgical and prognostic implications. SurgRadiol Anat 1993;15:271–277.

w9x Riquet M, Manac’h D, Dupont P, Dujon A, Hidden G, Debesse B. Anatomicbasis of lymphatic spread of lung carcinoma to the mediastinum:anatomo-clinical correlations. Surg Radiol Anat 1994;16:229–238.

w10x Rouviere H. Les vaisseaux lymphatiques des poumons et les ganglionsvisceraux intrathoraciques. Ann Anat Pathol 1929;6:113–158.

w11x Yano T, Yokoyama H, Inoue T, Asoh H, Tayama K, Ichinose Y. Surgicalresults and prognostic factors of pathologic N1 disease in non-small-cell carcinoma of the lung. Significance of N1 level: lobar or hilarnodes. J Thorac Cardiovasc Surg 1994;107:1398–1402.

w12x Van Velzen E, Snijder RJ, Brutel de la Riviere A, Elbers HRJ, van denBosch JMM. Lymph node type as a prognostic factor for survival inT2N1M0 non-small cell lung carcinoma. Ann Thorac Surg 1997;63:1436–1440.

w13x Riquet M, Manac’h D, Le Pimpec Barthes F, Dujon A, Chehab A.Prognostic significance of surgical-pathologic N1 disease in non-smallcell carcinoma of the lung. Ann Thorac Surg 1999;67:1572–1576.

w14x Munka V, Gregor A, Durovicova J. The topography and tributary areasof the bronchopulmonary lymph nodes of left lungs. Folia Morphol1972;20:201–203.

w15x Murray JF. The normal lung. The basis for diagnosis and treatment ofpulmonary disease. Philadelphia: WB Saunders Company; 1986.

w16x Mountain CF, Dresler CM. Regional lymph node classification for lungcancer staging. Chest 1997;111:1718–1723.

w17x Asamura H, Suzuki K, Kondo H, Tsuchiya R. Where is the boundarybetween N1 and N2 stations in lung cancer? Ann Thorac Surg 2000;70:1839–1846.

w18x Riquet M, Le Pimpec Barthes F, Lang-Lazdunski L. Boundary betweenN1 and N2 stations in lung cancer: back to the future of anatomy. AnnThorac Surg 2001;72:1439–1440.

w19x Rea F, Marulli G, Callegaro D, Zuin A, Gobbi T, Loy M, Sartori F.Prognostic significance of main bronchial lymph nodes involvement innon-small cell lung carcinoma: N1 or N2? Lung Cancer 2004;45:215–220.

w20x Okada M, Sakamoto T, Yuki T, Mimura T, Nitanda H, Miyoshi K, TsubotaN. Border between N1 and N2 stations in lung carcinoma: lessons fromlymph node metastatic patterns of lower lobe tumors. J Thorac Cardio-vasc Surg 2005;129:825–830.

w21x Lee JG, Lee CY, Park IK, Kim DJ, Cho SH, Kim KD, Chung KY. Theprognostic significance of multiple station N2 in patients with surgicallyresected stage IIIA N2 non-small cell lung cancer. J Korean Med Sci2008;23:604–608.

w22x Misthos P, Sepsas E, Kokotsakis J, Skottis I, Lioulias A. The significanceof one-station N2 disease in the prognosis of patients with non-small-cell lung cancer. Ann Thorac Surg 2008;86:1626–1631.

w23x Rusch V, Crowley J, Giroux DJ, Goldstraw P, Im JG, Tsuboi M, TsushiyaR, Vansteenkiste J. The IASLC lung cancer staging project: proposalsfor the revision of the N descriptors in the forthcoming seventh editionof the TNM classification for lung cancer. J Thorac Oncol 2007;2:603–612.

w24x Riquet M. Bronchial arteries and lymphatics of the lung. Thorac SurgClin 2007;17:619–638.

w25x Raymond DP, Daniel TM. Mediastinal anatomy and mediastinoscopy. In:Sellke FW, del Nido PJ, Swanson SJ, editors. Sabiston & Spencer surgeryof the chest, 7th edition. Philadelphia: Elsevier Saunders, 2005;657–666.

w26x Martin LW, Krasna MJ. Lymphatic drainage of the esophagus. In: ShieldsTW, Locicero III J, Reed CE, Feins RH, editors. General thoracic surgery,7th edition. Wolters KluweryLippincott Philadelphia: Williams & Wilkins,2009;1683–1689.

w27x Riquet M, Saab M, Le Pimpec Barthes F, Hidden G. Lymphatic drainageof the esophagus in the adult. Surg Radiol Anat 1993;15:209–211.

w28x Rusch VW, Asamura H, Watanabe H, Giroux DJ, Rami-Porta R, GoldstrawP. The IASLC lung cancer staging project. A proposal for a newinternational lymph node map in the forthcoming seventh edition ofthe TNM classification for lung cancer. J Thorac Oncol 2009;4:568–577.

w29x Edge SB, Byrd DR, Compton CC, Fritz AG, Greene FL, Trotti A. Lung.In: AJCC cancer staging hand book. From the AJCC Cancer StagingManual, 7th edition. Chicago: Springer, 2010;299–323.

Downloaded from https://academic.oup.com/icvts/article-abstract/11/3/260/813245by gueston 18 March 2018