肺叶、肺段淋巴结引流的解剖学特征

目的探讨肺叶、肺段淋巴结引流的解剖学特征。 方法对9具成人尸体采用解剖乳胶填充剂行胸部淋巴结灌注,然后游离标本的纵隔前、纵隔后及中纵隔淋巴结,同时游离并清扫右肺上、中、下肺叶和各个肺段,以及左肺上、下肺叶和各个肺段的肺内淋巴结、肺门淋巴结;观察淋巴结的分布、数目和淋巴回流状况。 结果在标本上共观察到212个纵隔淋巴结,平均每例23.5个;各区淋巴结的数目以隆突下淋巴结7区和右下气管旁4R最多,其次为右气管支气管旁(10R)、左支气管旁(10L)和主-肺动脉窗区(5区)淋巴结;纵隔各区以隆突下区(7区)淋巴结最大,其次是右气管支气管旁(10R)淋巴结,气管旁淋巴结自上而下直至隆突下淋巴结逐渐增大,并且右侧大于左侧,即下大于上,右大于左。左肺和右肺的肺内淋巴结一般按照亚段淋巴结→段淋巴结→叶淋巴结→叶间淋巴结/肺门淋巴结;右肺上叶、中叶及肺门淋巴结通常回流至上纵隔淋巴结及隆突下淋巴结,下叶回流至下纵隔淋巴结。而左肺上叶一般引流至主—肺动脉窗区淋巴结及隆突下淋巴结,下叶也引流至下纵隔淋巴结。 结论肺叶及纵隔淋巴回流具有一定的规律性,从而为肺叶特异性/系统性淋巴结清扫方式的选择提供了解剖学依据。     

Indication of limited pulmonary resection for small-sized lung cancer based on preoperative clinical data

Limited pulmonary resection is performed mostly based on the size of lung cancer and ground-glass opacity (GGO). It has been proposed to determine the indication of segmentectomy according to hilar lymph node involvement. There is a potential risk of underestimation for lymph node involvement since there may be a skip mediastinal lymph node metastasis without hilar involvement. We propose to use standardized uptake value( SUV) max of primary lung cancer as an indicator of non-invasive lung cancer. None of 44 small-sized lung cancers with SUVmax lower than 1 had lymph node metastasis or vessel invasion. A small-sized lung cancer ≤ 2 cm with SUVmax ≤ 1 is indicated wedge resection if GGO area is greater than 75% of tumor. Segmentectomy is indicated if the GGO area is less than 75%. We also propose selective lymphadenectomy for small-sized lung cancer. The lower mediastinal lymphadenectomy may be omitted if a small-sized tumor is located in the right upper lobe or the left upper segment. The upper mediastinal lymphadenectomy may be omitted if a small-sized lung cancer is located in the lower lobe and if the lower mediastinal lymph node involvement is excluded.     

直径≤3cm的周围型非小细胞肺癌纵隔淋巴结转移规律的初步研究

目的探讨直径≤3cm的周围型非小细胞肺癌(non-small cell lung cancer,NSCLC)纵隔淋巴结转移的情况,分析早期周围型NSCLC纵隔淋巴结转移的规律。方法 2000年1月1日～2008年12月31日治疗直径≤3cm的周围型NSCLC161例,男89例,女72例,年龄(63.4±10.7)岁,行肺叶切除或肺局限性切除加系统性纵隔淋巴结清扫术,分析其临床特征、病理特点及纵隔淋巴结转移规律。结果全组手术顺利,无死亡及严重并发症发生。肺叶切除153例,肺楔形切除7例,肺段切除1例。全组共清扫淋巴结2456枚,平均每例4.5±1.6组、13.1±7.3枚。术后病理:腺癌99例,鳞癌30例,肺泡细胞癌19例,其他类型肺癌13例。术后TNM分期:ⅠA期50例,ⅠB期62例,ⅡA期6例,ⅡB期10例,ⅢA期33例。N1组淋巴结转移率为23.6%(38/161),N2组转移率为20.5%(33/161),其中隆突下淋巴结转移率为8.1%(13/161),跳跃式纵隔转移率为6.8%(11/161),全组未发现下纵隔淋巴结转移。肺泡细胞癌及直径≤2cm的鳞癌、直径≤1cm的腺癌均无pN2转移。上肺癌发生pN2转移时上纵隔100%(19/19)受累,其中21.1%(4/19)同时伴有隆突下淋巴结转移;下肺癌则除主要转移至隆突下外(64.3%,9/14),还常直接单独转移至上纵隔(35.7%,5/14)。转移的纵隔淋巴结左肺癌主要分布在第5、6、7组,右肺癌主要分布在第3、4、7组。结论对于直径≤3cm的周围型NSCLC,肿瘤直径越大,其纵隔淋巴结转移率越高,肺泡细胞癌、直径≤2cm的鳞癌和≤1cm的腺癌其纵隔淋巴结转移率相对较低;上肺癌主要转移在上纵隔,下肺癌则隆突下及上纵隔均可转移;第5、6、7组淋巴结是左肺癌主要转移的位置,第3、4、7组是右肺癌主要转移的位置,术中应重点清扫。     

Rational lymph node dissection for lung cancer according to the occurrence lobe and histological type]

Between January 1989 and December 1998, 134 cases of squamous cell carcinoma and 244 cases of adenocarcinoma underwent surgical resection of the lung with systematic lymph node dissection in our hospital. The cN diagnosis by CT scan and pN diagnosis were compared. In squamous cell carcinoma pN 2-3 cases were only one patient (2%) out of 60 patients with cN 0, 5 patients (18%) out of 28 patients with cN 1, and 21 patients (46%) out of 46 patients with cN 2-3. On the other hand in adenocarcinoma pN 2-3 cases were 27 patients (14%) out of 193 patients with cN 0, 3 patients (25%) out of 12 patients with cN 1, and 24 patients (62%) out of 39 patients with cN 2-3. The pathways of the lymphatic metastases to the mediastinal nodes were analized in 27 patients with squamous cell carcinoma and 54 patients with adenocarcinoma undergoing systematic lymph node dissection. All patients had histologically proven mediastinal metastasis. Histologically there was no difference in pathways of the lymphatic metastases to the mediastinal nodes. 1. The dominant lymphatic drainage from the right upper lobe flowed into the superior mediastinal nodes. The direct metastatic passages to the superior mediastinal nodes were observed (47%). Subcarinal and inferior mediastinal node involvement was rare (3%). 2. The dominant lymphatic drainage from the middle and the lower lobe flowed into the subcarinal nodes (85%). The involvement of the superior mediastinal nodes occurred in 53% of subcarinal node positive patients on the right side. 3. The dominant lymphatic drainage from the left upper lobe flowed into the subaortic or paraaortic nodes (69%). Subcarinal and inferior mediastinal node involvement was rare (6%). We conclude that subcarinal and inferior mediastinal lymph node dissection is not necessary for upper lobe lung cancers, and that superior mediastinal lymph node dissection can be omitted in middle and lower lobe lung cancers without hilar and subcarinal lymph node involvement, especially in the cases of cN 0.     

Sentinel node sampling limits lymphadenectomy in stage I non-small cell lung cancer.

OBJECTIVE: It is controversial whether a systematic mediastinal lymph node dissection (MLND) needs to be performed in all patients with stage I lung cancer. The present study was done to examine the new sentinel lymph nodes hypothesis based on the lobe of the primary tumor. METHODS: In our first study, the lymph node (LN) metastases were assessed in 291 stage I non-small cell lung cancer (NSCLC) patients who had a major lung resection with a systematic mediastinal lymph node dissection. We evaluated the validity of using our new sentinel lymph nodes method based on the lobe of the primary tumor as follows: the pretracheal (#3), tracheobronchial (#4), and hilar nodes (#10) for right upper lobe tumors; #4, subcarinal (#7), and #10 for middle lobe tumors; the subaortic (#5), paraaortic (#6), and #10 for left upper lobe tumors; and the #7, #10, and interlobar nodes (#11) for tumors in either lower lobes. In the second study, we performed a lobectomy with new sentinel node sampling in 64 patients with preoperative complications. If all of the sampling nodes showed no metastases on frozen section diagnosis, systematic node dissections were not performed. RESULTS: Six of 291 patients in the first study had skip metastases that did not involve the new sentinel nodes; 5 of the 6 patients had macroscopic pleural invasion. Thus, we defined pleural invasion as an exclusion criterion for the second study. In the second study, the median follow-up time was 39 months. Metastatic lymph nodes were detected in 11 of 64 patients. Fifty-three patients (83%) had no metastasis in the sampled nodes, and, therefore, a mediastinal lymph node dissection was not done. The morbidity rate in the sampling group was 36%, and there was no mortality. In the sampling group, local recurrences were observed in two patients, distant metastases in eight, and carcinomatous pleuritis in one; the overall 5-year survival rate was 82%. CONCLUSIONS: We found that it is possible to perform a less invasive lymphadenectomy for patients with stage I lung cancer using intra-operative sampling of new sentinel lymph nodes.     

Mediastinal lymph node involvement as the initial manifestation of occult thyroid cancer in the surgical treatment of lung cancer: Report of a case

A 63-year-old man was referred to our institute for the treatment of squamous cell carcinoma of the upper lobe of his right lung. A right upper lobectomy of the lung was performed with a mediastinal lymph node dissection. The postoperative pathological examination of the dissected specimens revealed one of the superior mediastinal lymph nodes to be morbid with micrometastasis of occult thyroid cancer, while no node involvement was seen due to lung cancer. A right lobectomy of the thyroid gland with a modified radical neck dissection was done 4 years later after the confirmation of the absence of any recurrent sign of lung cancer. In the resected specimen, papillary thyroid microcarcinoma was observed with several intraglandular metastases and right regional lymph node involvement. Eight months later, a new primary lung cancer developed in the left lung, and a left upper lobectomy of the lung with a mediastinal lymph node dissection was performed. At that time, the absence of mediastinal lymph node metastasis from lung cancer or thyroid cancer was confirmed. Mediastinal lymph node involvement as the initial manifestation of occult thyroid cancer in surgical treatment for lung cancer is rare, but it is important to be aware of the possibility of incidentally detecting occult thyroid cancer in surgical dissections in this area for lung cancer. The appropriate surgical treatment should be determined while carefully considering the prognosis of the lung cancer as well as that of any coexisting malignancy.     

Limited mediastinal lymph node dissection for non-small cell lung cancer according to intraoperative histologic examinations

BACKGROUND: Although radioisotopic procedures are commonly used to detect sentinel lymph nodes in breast cancer surgery, these procedures are often problematic and not necessarily suitable for lung cancer surgery. METHODS: Our previous study revealed that the mediastinal sentinel lymph node, defined as the regional mediastinal lymph node, consisted of nodes 2, 3, or 4 in right upper lobe cancers; 3, 7, or 8 in right lower lobe cancers; 4, 5, or 7 in left upper lobe cancers; and 4, 7, or 8 in left lower lobe cancers. On the basis of these findings, we pathologically investigated one representative lymph node at each of the 3 levels dissected during surgical intervention in 69 patients with non-small cell lung cancer from September 1993 through December 2002. Fifty-eight patients with lung cancer underwent lobectomies with limited mediastinal lymph node dissection according to this strategy. RESULTS: Mediastinal lymph node recurrence was observed in only one patient during 41 +/- 25 months (maximum, 98 months) of follow-up. The cancer-specific 5-year survivals were 96.6% in patients with pathologic stage IA disease (n = 31) and 67.4% in patients with stage IB disease (n = 16). CONCLUSION: These results suggested that limited mediastinal lymph node dissection is applicable to patients with non-small cell lung cancer whose regional mediastinal lymph nodes are not metastatic.     

Lung cancer which accompanies anomalous venous connections, pulmonary and systemic-partial anomalous pulmonary venous connection and persistent left superior vena cava: report of 2 cases

We report the cases of a left partial anomalous pulmonary venous connection (PAPVC) and a persistent left superior vena cava (PLSVC), combined with primary lung cancer. Our case of PAPVC, the anomalous pulmonary vein originated from the hilum of the left upper lobe flowed into the left brachiocephalic vein. A left lower lobectomy was performed uneventfully without correcting the anomalous vein. And a case of PLSVC, the left superior vena cava flowed into the right superior vena cava, running under the aortic arch. A left upper lobectomy and mediastinal lymph node dissection was performed in safety. Although PLSVC was detected by chest computed tomography (CT) before operation, PAPVC was noticed intraoperatively in our case. We should keep in mind the possibility of variations of pulmonary vessel distribution, especially PAPVC located in a different lobe for resection, when undertaking lung resection.     

Omission of mediastinal lymph node dissection in lung cancer: its techniques and diagnostic procedures.

To reduce or omit a mediastinal lymph node dissection in the patients with clinical stage I non-small cell lung cancer (NSCLC), several authors examined the prevalence of metastatic sites of lymph nodes. Because lymphatic drainage usually heads for the upper mediastinum in upper lobe cancer and for the lower mediastinum in lower lobe cancer, upper and lower mediastinal lymph node dissection could be reduced in lung cancers of lower lobe and upper lobe. By using sentinel node (SN) navigation surgery, it is possible to omit mediastinal lymph node dissection. Radiological findings are also useful to determine reduction of mediastinal lymph node dissection. In clinical stage Ia adenocarcinomas that show ground glass opacity (GGO) findings on computed tomography (CT) or negative for fluorodeoxyglucose accumulation on positron emission tomography (PET), mediastinal lymph node dissection can be omitted, because these types of adenocarcinomas rarely metastasize to the lymph nodes. By using these procedures, mediastinal lymph node dissection can be reduced or omitted with little risk of local recurrence.     

Systemic mediastinal lymph node dissection of right lung cancer: surgical quality control and analysis of mediastinal lymph node metastatic patterns

Standardization of systemic mediastinal lymph node dissection (SMLD) of lung cancer requires further investigation. A consecutive 124 right lung cancer patients were recruited for pulmonary resection plus SMLD. Three mediastinal lymph node compartments, (i) the upper compartment (station 1-4), (ii) the middle compartment (station 7-8) and (iii) the lower compartment (station 9), were en bloc collected to achieve surgical quality control and to analyze mediastinal lymph node metastatic patterns. The number of total harvested lymph nodes, N1 nodes and N2 nodes were 21.9+/-8.7, 9.2+/-4.7 and 12.8+/-6.7, respectively. Tumor location (peripheral or central) (P=0.023) and status of blood vessel invasion (P=0.002) were identified as risk factors for nodal involvement. Right upper lobe (RUL) cancer with N2 disease primarily metastasized to the upper compartment (27.3%) (P=0.001). For right lower lobe (RLL) cancer, lymph node metastasis most commonly detected in the middle compartment (48.8%) (P=0.001). Single mediastinal compartment metastasis occurred in 64.7% (11/17) of adenocarcinomas from RUL and RML, whereas multiple compartments metastasis occurred in all adenocarcinoma cases (12/12) from RLL (P=0.001). SMLD needs to standardize the extent of lymphadenectomy and number of removed lymph nodes for surgical quality control. Simplifying mediastinal lymph node stations to three compartments may benefit surgical excision.     

Combined resection of the aorta for an esophageal carcinoma invading the aorta through a right transthoracic approach

The prognosis of esophageal carcinoma invading the thoracic aorta has been extremely poor, as it has been either not resected or only palliatively resected. In recent years a remarkable improvement in survival has been achieved in advanced esophageal carcinoma through an aggressive dissection of the upper mediastinal lymph nodes. This implied that resection only of the aorta without lymph node dissection in these patients was not adequate for curability. Although a resection of the aorta would seem to be performed more easily through a left thoracotomy than through a right thoracotomy, the upper mediastinal lymph node dissection was unsatisfactory through a left thoracotomy. Therefore, we performed combined resection of the aorta using a temporary aorta-aorta bypass together with upper mediastinal lymph node dissection through a right thoracotomy for four patients with the esophageal carcinoma invading the thoracic aorta. This operative procedure was performed safely, and had the advantage that full observation on the extent of the carcinoma was attained together with subsequent radical lymph node dissection in the same field through only the right thoracic approach. This operation may provide a possibility for cure to patients with an esophageal carcinoma invading the aorta, who would otherwise receive only palliative treatment.     

Combined resection of the aorta for an esophageal carcinoma invading the aorta through a right transthoracic approach.

The prognosis of esophageal carcinoma invading the thoracic aorta has been extremely poor, as it has been either not resected or only palliatively resected. In recent years a remarkable improvement in survival has been achieved in advanced esophageal carcinoma through an aggressive dissection of the upper mediastinal lymph nodes. This implied that resection only of the aorta without lymph node dissection in these patients was not adequate for curability. Although a resection of the aorta would seem to be performed more easily through a left thoracotomy than through a right thoracotomy, the upper mediastinal lymph node dissection was unsatisfactory through a left thoracotomy. Therefore, we performed combined resection of the aorta using a temporary aorta-aorta bypass together with upper mediastinal lymph node dissection through a right thoracotomy for four patients with the esophageal carcinoma invading the thoracic aorta. This operative procedure was performed safely, and had the advantage that full observation on the extent of the carcinoma was attained together with subsequent radical lymph node dissection in the same field through only the right thoracic approach. This operation may provide a possibility for cure to patients with an esophageal carcinoma invading the aorta, who would otherwise receive only palliative treatment.     

The right upper lobe pulmonary resection performed through the transcervical approach.

OBJECTIVE: Preliminary report: presentation of the new technique of transcervical right upper lobectomy with transcervical extended mediastinal lymphadenectomy (TEMLA) for NSCLC. METHODS: Two patients underwent the operation that was performed through the collar incision, with elevation of the sternal manubrium with the mechanical sternal retractor. TEMLA and bilateral mediastinal lymph node excision (stations 1, 2R, 4R, 2L, 4L, 3A, 3P, 7 and 8) and bilateral supraclavicular lymph node excision were performed (frozen section analysis: all nodes negative). The mediastinal pleura was opened and the following structures were dissected in the open fashion with standard surgical instruments and divided with the use of endostaplers: the azygos vein, the upper trunk of the right pulmonary artery, the branch of the superior pulmonary vein to the upper lobe, the upper lobe bronchus, the segment 2 artery, the posterior part of the oblique fissure and the horizontal fissure. The operation was performed with the use of one videothoracoscopic (VTS) port for insertion of 5mm, 30 degree VTS camera for intraoperative control and for single thoracic drain for the postoperative period. RESULTS: The operative times were 250 and 270 min, respectively; intraoperative blood loss was 110 and 100ml, respectively. There were no intraoperative complications. The postoperative course was remarkably smooth. The final pathologic report: large cell carcinoma pT2N0M0 and squamous cell carcinoma pT2N0M0, no metastatic changes of 51 and 41 mediastinal and intrapulmonary (stations 10, 11 and 12) and supraclavicular nodes, respectively. CONCLUSIONS: This preliminary report indicates possible advantages of the transcervical right upper lobe pulmonary resection including: (1) extremely radical, minimal invasive procedure with no need for utility thoracotomy; (2) dissection performed with standard surgical instruments in the open fashion.     

Long-term survival cases of lung cancer presented as solitary bone metastasis.

The prognosis of non-small cell lung carcinoma (NSCLC) with bone metastasis has been regarded as very poor. We report herein on two cases of NSCLC which presented as a solitary bone metastasis, were treated with surgical resection. Both these cases survived for over 5 years after their last operations. A 71-year-old-man was hospitalized with right crural pain. A diagnosis of squamous cell carcinoma of the left lower lobe with right fibula metastasis was made. A marginal resection of the right fibula was performed. After that, a left lower lobe lobectomy and systemic chemotherapy were carried out. He had a local recurrence in the right mediastinal lymph nodes eleven months after the operation. He received intraluminal and external radiation therapy and obtained complete remission. He has survived for 5 years without any other recurrence or metastasis. A 52-year-old-man was admitted to our hospital with left thigh pain. A diagnosis of adenocarcinoma of the right upper lobe with left thigh metastasis was made. A right upper lobe lobectomy and a resection of the left thigh tumor were performed. Three cycles of systemic chemotherapy were given after that. He has survived for 5 years since his last operation without any recurrence or metastasis.     

The complete hilar-mediastinal lymph node dissection

Starting from Cahan's "radical pneumonectomy" and "radical lobectomy", mediastinal lymph node dissection was introduced in Japan by Ishikawa and survival results analyzed by Naruke. Japanese Lung Cancer Society (JLCS) introduced Naruke's lymph node map to standardizing dissection. Upper mediastinum, subcarinal, interlobar and upper lobar lymph nodes are to be dissected for tumors located in the right upper lobe, and the same areas in the case of middle lobe. In tumors of the lower right lobe, also nodes of the lower mediastinum should be dissected. When the tumor is in the left upper lobe, upper mediastinum (except pre and paratracheal lymph nodes), subcarinal, interlobar and lobar nodes should be dissected. Finally, in left lower tumors, lower mediastinum is to be dissected. Super-radical dissection is performed through a median sternotomy to reach pre and paratracheal nodes in tumors affecting the left upper lobe.     

Thoracoscopic management including lobectomy for a primary peripheral pulmonary adenoid cystic carcinoma in a cirrhotic patient

Primary peripheral adenoid cystic carcinoma is a relatively rare type of lung carcinoma. A 49-year-old male with significant hepatic cirrhosis presented with an incidental right upper lobe nodule suspicious for malignancy and an additional imaging suggesting contralateral mediastinal lymph node involvement. Thoracoscopic management included a left-sided lymph node dissection and a right-sided lobectomy in this medically high-risk patient.     

Video-assisted thoracic surgery for left lung cancer in a patient with a right aortic arch

We report a rare case of left lung cancer in a patient with a right aortic arch. A 65-year-old woman was diagnosed to have an adenocarcinoma in the left upper lobe (S3) in addition to a right aortic arch (type II), with the left subclavian artery originating from the descending aorta. Left upper lobectomy and lymph node dissection was performed by video-assisted thoracic surgery (VATS). For the mediastinal dissection, the upper mediastinal lymph nodes were easily resected after verifying the location of the arterial ligament and the recurrent laryngeal nerve (RLN). This is the first report of using VATS to remove a lung cancer from a patient with a right aortic arch.     

Double primary lung carcinoma consisting of large cell neuroendocrine carcinoma and squamous cell carcinoma: report of a case.

We report a rare case of double primary lung carcinoma including large cell neuroendocrine carcinoma (LCNEC). A 67-year-old man underwent an annual medical checkup in 2000, pulmonary carcinoma was strongly suspected by sputum cytology and radiological images. Preoperative diagnosis was double primary lung carcinoma with a squamous cell carcinoma in the right lower lobe and non-small cell carcinoma in the right upper lobe. The histological carcinoma type in the right upper lobe could not be determined preoperatively. The patient underwent a right lower lobectomy and wedge resection of the right upper lobe. Histologically, the tumor in the right upper lobe was LCNEC and the tumor in the right lower lobe was a moderately differentiated squamous cell carcinoma. The patient had right supraclavicular lymph node metastases of LCNEC and died of multiple pulmonary metastases 10 months after the operation.     

In what case can we abbreviate mediastinal lymph nodes dissection of peripheral small-sized lung cancer?

We analyzed 723 cases of non small cell lung cancer (459 adenocarcinomas and 264 squamous cell carcinomas) from the view point of lymph nodes metastases, according to histological type, location of cancer and tumor size. METHOD: Histological type was adenocarcinoma or squamous cell carcinoma. Location was divided into 8 areas [right side; 4 areas, upper lobe (RUL)/middle lobe (RML)/S6 (RS6)/basal segment of lower lobe (RBS): left side; 4 areas, upper division of upper lobe (LUD)/lingula (LLS)/S6 (LS6)/Basal segment of lower lobe (LBS)]. Tumor size was divided by centimeters, namely 0.0-1.0 cm, 1.1-2.0 cm, 2.1-3.0 cm, etc.. RESULTS: Safety size of lung cancer in which we can abbreviate mediastinal lymph nodes dissection was as follows. In adenocarcinoma, in RUL/RML/RBS 1.0 cm, in RS6 2.0 cm, in LUD 1.0 cm, in LLS/LS6/LBS 2.0 cm. In squamous cell carcinoma, in RUL 1.0 cm, in RML/RS6/RBS 2.0 cm, in LUD 1.0 cm, in LLS/LS6/LBS 2.0 cm. CONCLUSION: In 1.0 cm or smaller non small cell lung cancer we might abbreviate mediastinal lymph nodes dissection. Moreover, in squamous cell carcinoma of (RML, LLS, right or left lower lobe) of 2.0 cm or smaller size, we might abbreviate mediastinal lymph nodes dissection.     

Results of surgical intervention for p-stage IIIA (N2) non-small cell lung cancer: acceptable prognosis predicted by complete resection in patients with single N2 disease with primary tumor in the upper lobe

OBJECTIVES: Non-small cell lung cancer with mediastinal lymph node involvement is a heterogeneous entity different from single mediastinal lymph node metastasis to multiple nodes or extranodal disease. The objective of this study was to identify the subpopulation of patients with N2 disease who can benefit from surgical intervention. METHODS: We reviewed 219 consecutive patients with N2 non-small cell lung cancer treated with a thoracotomy between November 1980 and June 2002 and retrospectively analyzed 154 of those who had p-stage IIIA disease and underwent a complete resection. Age, sex, side (right or left), histology, location (upper or middle-lower lobe), tumor size, c-N factor, and N2 level (single or multiple) were used as prognostic variables. RESULTS: The 3- and 5-year survivals were 45.3% and 28.1%, respectively, in patients with p-stage IIIA (N2) disease. Survival for those with single N2 non-small cell lung cancer was significantly better than in those with multiple N2 disease (P =.0001), and patients with a tumor in the upper lobe showed a significantly longer survival than those with middle-lower lobe involvement (P =.0467). The 3- and 5-year survivals for patients with single N2 disease with a primary tumor in the upper lobe were 74.9% and 53.5%, respectively. A multivariate analysis with Cox regression identified 5 predictors of better prognosis: younger age, squamous cell carcinoma as determined by histology, primary tumor location in the upper lobe, c-N0 status, and a single station of mediastinal node metastasis. CONCLUSION: Our results suggest that of the heterogeneity of N2 diseases, patients with single N2 disease with non-small cell lung cancer in the upper lobe are good candidates for pulmonary resection.