| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
J Thorac Cardiovasc Surg 1994;108:321-328
© 1994 Mosby, Inc.
GENERAL THORACIC SURGERY |
Kyoto, Japan
From the Respiratory Division, Kyoto-Katsura Hospital, Kyoto, Japan.
Received for publication Sept. 14, 1993. Accepted for publication Jan. 9, 1994. Address for reprints: T. Funatsu, MD, Respiratory Division, Kyoto- Katsura Hospital, 17, Yamadahirao, Nishikyo-ku, Kyoto (615), Japan.
Abstract
The significance of preoperative N factor assessment in T1 lung cancer and the need for mediastinal node dissection in T1 N0 M0 cases were investigated. The results of mediastinoscopy were evaluated in patients with T1 adenocarcinoma or squamous cell carcinoma who underwent preoperative mediastinoscopy and thoracotomy from 1971 to 1991 (n = 164). Mediastinoscopy gave true negative results in 90% of patients, false negative in 1%, and true positive in 9%. The 5-year survival was 90% for patients with T1 N0 M0 disease who underwent nonradical dissection (n = 64) and 70% for those who underwent radical dissection (n = 61), indicating that the prognosis was significantly better (p < 0.05) with nonradical dissection. Distant metastasis was a common cause of death, and no death was related to local recurrence, whether nonradical or radical dissection had been performed. The results strongly suggest that preoperative mediastinoscopy and intraoperative node sampling are sufficient for assessment of N factors in T1 lung cancer. The possible relation between immunologic functional changes associated with mediastinal lymph node dissection and the prognosis in patients without evidence of positive lymph nodes should be clarified by a prospective randomized study. (J THORAC CARDIOVASC SURG 1994;108:321-8)
The standard operative procedures in many hospitals and institutions still include mediastinal lymph node dissection even for patients with T1 N0 M0 lung cancer without nodal involvement. 
1-3 One reason is that N factors cannot be accurately assessed by diagnostic imaging before the operation. The underlying idea is to simultaneously make an accurate diagnosis of N factors and dissect metastatic lymph nodes. 1-3 In contrast, much evidence exists 4-10 that N factors can be accurately assessed before the operation if mediastinoscopy is used alone or together with computed tomographic scanning. In addition, virtually no prospective studies investigating the long-term effect of mediastinal node dissection in node-negative cases have been conducted. Mediastinal node involvement is relatively uncommon in T1 lung cancer. Therefore it is important to investigate the accuracy of preoperative N factor assessment in relation to the need for mediastinal node dissection in this type of disease. The present article presents the results of mediastinoscopy, which was used to assess N factors before surgery, and compares retrospectively the long-term prognosis of T1 N0 M0 disease after resection between patients who also underwent mediastinal node dissection and those who did not.
SUBJECTS AND METHODS
From April 1971 to December 1991, 181 patients with T1 lung cancer, adenocarcinoma, or squamous cell carcinoma, without evidence of distant metastasis, underwent thoracotomy. The results of mediastinoscopy and N factor data were analyzed in 164 of these patients (91%) who had undergone preoperative mediastinoscopy, according to histologic type, location, and size of tumors. Single lobes were resected in all these patients except three who had two lobes resected. Of the remaining 17, four underwent partial resection because their performance status was not good enough to tolerate total resection and 13 underwent lobectomy because lung cancer was diagnosed after exploratory thoracotomy. They did not undergo preoperative mediastinoscopy and so were excluded from analysis. To evaluate whether mediastinal node dissection was needed in patients with T1 lung cancer without positive nodes, we further analyzed the 5-year survivals and deaths within 5 years of 125 patients with N0 disease (of the 164 with T1 disease) according to histologic type, location, and size of tumors. The remaining 39 (N1 = 16, N2 = 20, N3 = 3) with positive nodes were excluded from evaluation of the prognosis.
Biopsy specimens were obtained of five lymph nodes (right and left paratracheal nodes [Nos. 2R and 2L]), tracheobronchial nodes (Nos. 4R and 4L), and a node at the tracheal bifurcation (No. 7) during mediastinoscopy. Immediately after biopsy, specimens were frozen in a cryostat, and pathologic examinations for metastases were done.
Thoracotomy was performed after mediastinoscopy on the same day. If the results of the mediastinoscopic examination were negative, lymph nodes at the site of mediastinoscopic biopsy were sampled again during the operation for a second pathologic examination. Because mediastinoscopic biopsy is anatomically difficult for all lymph nodes at the tracheal bifurcation, biopsy tissue was taken again in all patients to confirm whether metastases were present. Mediastinal nodes that could not be reached by the mediastinoscope (Nos. 5, 6, 8, and 9) were also sampled to examine for possible metastasis. The lymph nodes sampled during mediastinoscopy, all hilar and mediastinal nodes that were sampled during the operation, and all nodes in the resected lungs were embedded in paraffin after the operation for further histopathologic examination.
True node-negative cases (TN) were defined as those negative both mediastinoscopically and on postoperative biopsy. TN included N0 cases without any node involvement, N1 cases with a positive intrapulmonary or hilar node, and N2 cases with positive mediastinal nodes, which could not be reached by the mediastinoscope. False node-negative cases (FN) were defined as those with intraoperative or postoperative positive findings in any node that could be reached by the mediastinoscope, and accordingly mediastinoscopic biopsy did not seem to be accurate enough. True node-positive cases (TP) were defined as N2 or N3 cases with mediastinoscopic evidence of node involvement. The results of mediastinoscopic examination were assessed by Galen's method. 5,6,11 Sensitivity was calculated as TP/(TP + FN), accuracy as (TP + TN)/(TP + TN + FN), and negative predictive value as TN/(TN + FN).
Of patients with T1 N0 disease (n = 125), those who did not undergo mediastinal node dissection and those who underwent only resection of regional hilar nodes with mediastinal node sampling were regarded as having undergone nonradical dissection (n = 64). Radical dissection (n = 61) was defined as complete removal of superior and inferior mediastinal nodes ipsilateral to the tumor together with nodes at the tracheal bifurcation. Primary lesions in the right (n = 69) and left (n = 56) lobes were separately analyzed. Resected tumors were classified by size into those 2 cm or less in largest diameter (n = 72) and those between 2 cm and 3 cm (n = 53).
Deaths of patients with T1 N0 disease within 5 years were analyzed by cases, ipsilateral local recurrence in the hilum, mediastinum, or thoracic cavity, or distant metastasis. In addition to these causes of death, two types of cancer and other diseases were also considered. The frequency of these causes was analyzed in relation to whether radical or nonradical mediastinal node dissection has been performed.
Prognosis was analyzed on the basis of the results up to December 1992. Surgical deaths within 1 month after operation and those caused by other diseases were also included in the analysis. Survivals were calculated by the Kaplan-Meier method, and the difference between the two groups was analyzed for significance by the Cox-Mantel test. The results of mediastinoscopy and data on N factors in T1 cases and mortality data for different causes of death in T1 N0 cases were analyzed by the t test. A p value less than 0.05 was defined as being statistically significant.
RESULTS
As shown in Table I, involvement of superior mediastinal nodes that was assessable by mediastinoscopy was relatively uncommon in T1 cases, with TN at 90%, FN at 1%, and TP at 9%. FN was found only in two patients with adenocarcinoma in the right upper lobe. Metastasis was detected in nodes 2R and 4R. The sensitivity of mediastinoscopy in T1 cases was 88%, accuracy 99%, and negative predictive value 99%. Mediastinoscopic TN, FN, and TP were found to be 89%, 2%, and 9%, respectively, for adenocarcinoma (n = 122), and 93%, 0%, and 7%, respectively, for squamous cell carcinoma (n = 42). No significant difference was detected between the two histologic types. The corresponding values for right lobe cancer (n = 90) were 87%, 2%, and 11% and those for left lobe cancer (n = 74) were 95%, 0%, and 5%. This difference between the two different sites again was not significant. The tumors were classified according to size as small, 2 cm or less (n = 84) in diameter, and large, between 2 and 3 cm (n = 80). Mediastinoscopic TN, FN, and TP were 96%, 0%, and 4% for the small tumors and 84%, 2% and 14% for the larger tumors. This difference between the large and small tumors was significant for TN and TP.
|
Table II shows data by N factors on patients with T1 disease who underwent mediastinoscopy and thoracotomy. N0 disease was found in 125 patients (76%), N1 in 16 (10%), N2 in 20 (12%), and N3 in 3 (2%). Of patients with N2 disease, N2-0 (with metastasis in nodes that could not be reached by the mediastinoscope) was found in 7 (5%). Among these 7, node 3a was positive in 1 of the 3 with right lobe cancer and node 8 was positive in the remaining 2; node 5 was positive in all 4 with left lobe cancer. Contralateral mediastinal node involvement was found in 2 patients with adenocarcinoma in the right lung and in 1 with squamous cell carcinoma in the left lung. The prevalence of N0 and N2 disease differed significantly between small tumors of 2 cm or less and large ones of between 2 and 3 cm. N2 and N3 diseases were uncommon (6% and 1%) for small lesions of 2 cm or less.
|
Table III shows deaths within 5 years in 125 patients with T1 N0 M0 disease classified separately by histologic type, tumor location, and tumor size for those who underwent radical lymph node dissection and those who underwent nonradical dissection. The percentages for patients who underwent nonradical dissection and for those who underwent radical dissection virtually did not differ by histologic type, tumor location, or tumor size. Deaths within 5 years differed significantly between the two groups, with 5 deaths (8%) among patients having nonradical dissection and 13 (21%) among those having radical dissection. The difference was particularly marked for adenocarcinoma, with 10 (22%) deaths among patients having radical dissection and 3 (7%) among those having nonradical dissection. A common cause of death in the 125 patients was distant metastases, whether they underwent nonradical or radical dissection, and no death was related to local recurrence. Ten deaths (16%) resulted from distant metastases after radical dissection and 3 (5%) after nonradical dissection, indicating a significant difference (p < 0.05) between the two surgical procedures. Fig. 1 shows the 5-year survival in patients with T1 N0 M0 disease who underwent nonradical or radical dissection. It was 81% for all patients with T1 N0 M0 disease combined. When analyzed in relation to lymph node dissection, 5-year survival was 70% for those who underwent radical dissection and 90% for those who underwent nonradical dissection. It was significantly lower with radical dissection than with nonradical dissection. The 5-year survival was 81% for patients with adenocarcinoma and 84% for those with squamous cell carcinoma, indicating no significant difference between the two histologic types. For patients with adenocarcinoma, it was 64% if radical dissection was performed and 92% if nonradical dissection was performed, indicating a significant difference between the two modalities (Fig. 2). For patients with squamous cell carcinoma, it was 80% and 87%, without any significant difference between radical and nonradical dissection.
|
|
|
DISCUSSION
We have performed preoperative mediastinoscopy in almost all patients who were eligible for surgical operation, to assess N factors before the operation. 6
Many reports have presented the results of mediastinoscopy in patients with lung cancer, 4-10,12 whereas detailed analysis of the results of mediastinoscopy by N factors, in particular those of T1 cases, has been limited. 7,9 The percentage of mediastinoscopic positive nodes was 14.9% (n = 202) in T1 cases according to Coughlin and associates 7 and 10% (n = 145) according to Jolly and colleagues. 9 These percentages are a little higher than that of 9% in our report. The likely reason is the difference in patient population. Our study included only patients with adenocarcinoma or squamous cell carcinoma, whereas the other investigators analyzed patients with other types of cancers as well. Adenocarcinoma is more likely to be node positive than squamous cell carcinoma. 9 However, analysis of the percentage of patients with positive nodes with different T factors in relation to the histologic classification of the tumors has been limited. Our data on T1 cases showed that the percentage of positive nodes was 9% for patients with adenocarcinoma and 7% for those with squamous cell carcinoma, indicating no major difference between the two histologic types. Coughlin and coworkers 7 analyzed mediastinoscopic data by the location of tumors, although they did not consider T factors. They reported that 32.65% of right lobe lesions were node positive, which was significantly higher than the level of 13.73% for left lobe lesions. Our data on T1 cases similarly showed a higher TP rate in right lobe lesions. However, the reason for the lower rate in left lobe lesions was that a biopsy specimen of lymph node 5 in the left lobe could not be obtained with a mediastinoscope. As shown in 
Table II, in all four patients with left lobe lesions of N2-0, node 5 was positive. If these four patients were added to those with TP, the percentage of node-positive cases would not differ significantly between right and left lobe lesions. Analysis of mediastinoscopic results by the size of the tumors showed that TP was significantly less common with lesions 2 cm or less in diameter than with larger lesions between 2 and 3 cm. A difference in staging might be the reason for this discrepancy.
Sensitivity and accuracy are used to assess the performance of mediastinoscopy. 5-7 However, virtually no reports have been published on the analysis of these parameters by T factors. In our series, because FN was uncommon, sensitivity, accuracy, and negative predictive value were high at 88%, 99%, and 99%, respectively. Mediastinoscopy showed no FN case in T1 cases of squamous cell carcinoma, in left lung cancers, or in tumors 2 cm or less in diameter.
Many investigators 13-15 have reported on the value of computed tomography and other forms of diagnostic imaging in preoperative assessment of N factors. However, Izbicki and associates 16 carried out a prospective study in 108 patients with lung cancer and found that computed tomography was not accurate in diagnosis of lymph node involvement. In practice, we use computed tomography to identify and evaluate the location and size of mediastinal lymph nodes, and we apply the resulting information to the assessment of N factors, but we do not depend on it in deciding whether an operation should be performed or the extent of node dissection. 6 Mediastinoscopy, which is invasive and can give the histologic diagnosis, may be used together with computed tomography, which cannot give histological diagnosis, 9,17 but they should not be directly compared with each other.
The most important aspect of assessment of N factors in lung cancer is to confirm with histologic evidence whether mediastinal lymph nodes are involved. The only way to know this accurately is by preoperative lymph node biopsy with a mediastinoscope or by thoracotomy. 18 If preoperative mediastinoscopy is not performed, mediastinal node dissection or sampling, which also enables intraoperative lymph node biopsy, is indispensable. Surgeons who advocate mediastinal node dissection emphasize that it is a necessary preventive measure because of the possibility of skipping metastases or micrometastases in lymph nodes and for accurate assessment of N factors. 1,2,19 The percentage of positive mediastinal nodes has been reported to be 10% to 21% in T1 cases, but these levels are much lower than those in T2 cases. 7,9,14,15 Our data suggest that mediastinal node dissection was indicated in only 14% of patients with T1 lesions, 12% of those with N2, 2% of those with N3, and, in particular, only in 7% of those with small lesions 2 cm or less in diameter (see Table II). Jolly and associates 9 similarly reported that N2 lesions were found only in 5.9% of tumors 2 cm or less in size (n = 68). On the other hand, many investigators believe that the risk of mediastinal node dissection is negligible if based on short-term data immediately after surgery. 1-3 Few studies have been published that investigated the effects on long-term prognosis of mediastinal node dissection, which extensively resects normal mediastinal nodes and the lymphatic network.
The present article retrospectively analyzes the prognosis of T1 N0 M0 disease, confirmed by preoperative mediastinoscopy and intraoperative and postoperative studies, separately for those patients who underwent radical dissection of mediastinal nodes and those who underwent nonradical dissection. Radical and nonradical dissection were chosen according to the judgment of the 15 surgeons involved. About half of them used mediastinal node dissection in case of possible micrometastasis, and the other half were critical of it because they emphasized the importance of preserving the immunologic function of lymph nodes. The possibility that patients who underwent nonradical dissection might be left with positive nodes could not be denied. However, that no deaths occurred as a result of local recurrence within 5 years suggests that no positive node may have remained.
Many investigators have reported on the 5-year survival of patients with T1 N0 M0 disease. 18,20-28 Mountain 26 and Naruke and associates, 27 who performed mediastinal node dissection without preoperative mediastinoscopic evidence, reported 5-year survivals of 68.5% and 76.4%, respectively. In contrast, Little and coworkers, 28 who did not perform mediastinal node dissection although they assessed N factors by mediastinal node sampling during thoracotomy, found a 5-year survival of 72% in patients with T1 N0 M0 disease. We assessed N factors by preoperative mediastinoscopy and intraoperative node sampling. Our 5-year survival was 81%, which was similar to those presented by other investigators. What was important in our data was that the 5-year survival in patients who underwent nonradical dissection was as high as 90%, which exceeded all levels previously reported. Our data were also analyzed according to the histologic type of the tumors. Five-year survival did not differ significantly in patients with squamous cell carcinoma, whether radical or nonradical dissection was used. On the other hand, it was 64% in those with adenocarcinoma who underwent radical dissection, a significantly worse result than the 92% survival in those who underwent nonradical dissection. The reason is not known. Deaths caused by distant metastasis were significantly more common in patients who underwent radical dissection than in those who underwent nonradical dissection. Published data on postoperative recurrence in patients with T1 N0 M0 disease have shown that distant metastases to the brain and bones and double cancers are common, whereas local recurrence is not. 21-23
Whether mediastinal node dissection is good or bad in lung cancer, is both an old and a new question, which has not yet been answered. Many surgeons believe that primary lobectomy plus mediastinal node dissection are the standard procedures for lung cancer. 1-3 These procedures are recommended for advanced cancer with positive lymph nodes, but it is questionable whether they are also good for T1 N0 M0 disease, which is often at an early stage. Diagnostic techniques have advanced, and more accurate staging is becoming a reality. In addition, lymph nodes are important for their immunologic function. 29-32 Now is the time to review data and select the right patients for mediastinal node dissection, which should be used at the right stage of the disease. The present report is based on a retrospective review. We strongly recommend that a prospective randomized study be conducted, taking into consideration T and N factors and other background data of the patients. The results of our study showed a better prognosis in T1 N0 M0 disease for those patients who had not undergone radical dissection. They strongly suggest the need to determine whether the current standard procedures, which include complete mediastinal dissection, should be used in all patients with lung cancer, in particular in those with early cancer.
References
This article has been cited by other articles:
|
|
 |
|
  F. J. F. Herth, R. Eberhardt, M. Krasnik, and A. Ernst Endobronchial Ultrasound-Guided Transbronchial Needle Aspiration of Lymph Nodes in the Radiologically and Positron Emission Tomography-Normal Mediastinum in Patients With Lung Cancer Chest, April 1, 2008; 133(4): 887 - 891. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 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]
|
|
|
|
|
|
R. F. Kelly, T. Tran, A. Holmstrom, J. Murar, and R. J. Segurola Jr Accuracy and Cost-Effectiveness of [18F]-2-Fluoro-Deoxy-D-Glucose-Positron Emission Tomography Scan in Potentially Resectable Non-small Cell Lung Cancer Chest, April 1, 2004; 125(4): 1413 - 1423. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Leschber, G. Holinka, and A. Linder Video-assisted mediastinoscopic lymphadenectomy (VAMLA) - a method for systematic mediastinal lymphnode dissection Eur. J. Cardiothorac. Surg., August 1, 2003; 24(2): 192 - 195. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. Venissac, M. Alifano, and J. Mouroux Video-assisted mediastinoscopy: experience from 240 consecutive cases Ann. Thorac. Surg., July 1, 2003; 76(1): 208 - 212. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. S. Choi, Y. M. Shim, J. Kim, and K. Kim Mediastinoscopy in patients with clinical stage I non-small cell lung cancer Ann. Thorac. Surg., February 1, 2003; 75(2): 364 - 366. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. J. Lewis, R. J. Caccavale, J.-P. Bocage, and M. D. Widmann Video-Assisted Thoracic Surgical Non-Rib Spreading Simultaneously Stapled Lobectomy: A More Patient-Friendly Oncologic Resection Chest, October 1, 1999; 116(4): 1119 - 1124. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. J Lewis Is radical lymphadenectomy a valid oncologic procedure? Eur. J. Cardiothorac. Surg., September 1, 1999; 16(suppl_1): S11 - S12. [Full Text] [PDF]
|
|
|
|
 |
|
 K. J. M. Chan, MB, U. Pastorino, MD, and P. Goldstraw, FRCS SYSTEMATIC NODAL DISSECTION IN THE INTRATHORACIC STAGING OF PATIENTS WITH NON-SMALL CELL LUNG CANCER J. Thorac. Cardiovasc. Surg., February 1, 1999; 117(2): 246 - 251. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Maruyama, K. Sugio, T. Mitsudomi, G. Saitoh, T. Ishida, and K. Sugimachi RELATIONSHIP BETWEEN EARLY RECURRENCE AND MICROMETASTASES IN THE LYMPH NODES OF PATIENTS WITH STAGE I NON-SMALL-CELL LUNG CANCER J. Thorac. Cardiovasc. Surg., October 1, 1997; 114(4): 535 - 543. [Abstract] [Full Text]
|
|
|
|
|
|
R. J. Lewis, R. J. Caccavale, G. E. Sisler, J.-P. Bocage, and J. W. Mackenzie One Hundred Video-Assisted Thoracic Surgical Simultaneously Stapled Lobectomies Without Rib Spreading Ann. Thorac. Surg., May 1, 1997; 63(5): 1415 - 1421. [Abstract] [Full Text]
|
|
|
|
|
|
T. Takizawa, M. Terashima, T. Koike, H. Akamatsu, Y. Kurita, and A. Yokoyama MEDIASTINAL LYMPH NODE METASTASIS IN PATIENTS WITH CLINICAL STAGE I PERIPHERAL NON-SMALL-CELL LUNG CANCER J. Thorac. Cardiovasc. Surg., February 1, 1997; 113(2): 248 - 252. [Abstract] [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |
