| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 1997;63:1441-1450
© 1997 The Society of Thoracic Surgeons
Departments of Pulmonology (Respiratory Tumor Unit) and Thoracic Surgery, University Hospital Gasthuisberg, Catholic University Leuven, Belgium
Accepted for publication December 13, 1996.
 |
Abstract |
|---|
 TopFootnotes Abstract IntroductionPatients and MethodsMethodsStatistical AnalysisResultsUnivariate Analysis of...Multivariate AnalysisCommentAppendix 1AcknowledgmentsReferences |
|---|
Methods. One hundred forty patients with resected non–small cell lung cancer who eventually proved to have pathologic N2 disease were studied with a univariate and multivariate analysis of prognostic factors.
Results. Nineteen patients had a positive mediastinoscopy; the others had a preoperative N0 or N1 stage. Complete resection rate was 80.7%. Five-year survival was 20.8% (95% confidence interval, 17.2% to 24.4%), 32.2% in mediastinoscopy-negative patients. In the univariate analysis, clinical N stage at mediastinoscopy, complete resection, performance status, T stage, number of metastatic levels in adenocarcinoma, and nodal capsule rupture were important factors. In a multivariate model, survival was worse in case of higher T stage (relative risk = 1.43), lower performance status (relative risk = 1.37), involvement of more than one node level (relative risk = 1.68), nonsquamous histology (relative risk = 1.29) and clinical N2 stage (relative risk = 1.43). Long-term survival was unlikely when lactic dehydrogenase or carcinoembryonic antigen levels were elevated.
Conclusions. In clinical N0 or N1 cancer, complete resection resulted in reasonable survival prospects. In patients with N2 disease discovered at mediastinoscopy, surgical treatment was only worthwhile in case of minimal N2. Several unfavorable prognostic factors could be identified in the univariate analysis and confirmed in a multivariate Cox model.
|
Introduction |
|---|
|
TopFootnotesAbstractIntroductionPatients and MethodsMethodsStatistical AnalysisResultsUnivariate Analysis of...Multivariate AnalysisCommentAppendix 1AcknowledgmentsReferences |
|---|
In the early stages of non–small cell lung cancer (NSCLC), either without lymph node metastases (stage I) or with metastases to intrapulmonary or hilar lymph nodes only (stage II), complete resection offers a good prognosis with an expected 5-year survival of 50%, even 70% in the T1 N0 subset.
Patients with locally advanced disease (stage III) are a heterogeneous group. They are divided into stage IIIB (usually not considered for surgical treatment) and stage IIIA (potentially operable). Even stage IIIA has a variable prognosis: patients with a T3 N0 or T3 N1 tumor have a far better prognosis than those with stage IIIA N2 disease, especially if the T3 factor is caused by resectable chest wall invasion.
There is controversy in the literature about the indication for primary surgical treatment in stage IIIA N2 patients. In the past, some authors have stressed that the preoperative discovery of N2 disease is an ominous finding [1–3]. Others have advocated that, even in clinical N2 (cN2) disease, surgical exploration should always be undertaken to obtain 5-year survival rates as high as 20% [4].
This retrospective study reports the survival of 140 NSCLC patients with resected N2 disease. By means of a univariate and multivariate analysis, it also seeks the prognostic factors determining the survival of these patients.
|
Patients and Methods |
|---|
|
TopFootnotesAbstractIntroductionPatients and MethodsMethodsStatistical AnalysisResultsUnivariate Analysis of...Multivariate AnalysisCommentAppendix 1AcknowledgmentsReferences |
|---|
Pretreatment staging consisted of clinical history and physical examination, blood tests including a complete blood count, measurement of serum calcium level, liver function tests, and in some cases measurement of carcinoembryonic antigen level. All patients had a preoperative bronchoscopy and computed tomographic (CT) scan of the mediastinum and lungs. Lymph nodes with a long-axis diameter of 15 mm or more were considered abnormal. Upper abdominal ultrasonography was performed in all patients, completed by upper abdominal CT scan in case of equivocal findings. Brain CT scan was requested in case of neurologic symptoms or signs, and in all patients with adenocarcinoma or large cell undifferentiated carcinoma. Bone scintigraphy was carried out in patients with bone pain or raised alkaline phosphatase or serum calcium levels, completed by bone radiographs or a bone CT scan in case of equivocal findings. All patients with a large cell undifferentiated carcinoma or adenocarcinoma underwent surgical mediastinal exploration. In patients with squamous cell tumors, this evaluation was restricted to those with enlarged MLN on CT scan, or with a more advanced clinical T stage (T3 or T4). Surgical mediastinal exploration consisted of cervical mediastinoscopy, completed by left anterior mediastinotomy in appropriate cases. Labeling of MLN was performed according to the classification of Naruke and associates [5]. In 121 patients, the preoperative findings suggested clinical N0 or N1 disease and the presence of N2 was discovered at thoracotomy. In case of a positive mediastinoscopy, surgical treatment was attempted only in very select cases with younger age and good general condition.
The indication for postoperative radiotherapy was left at the discretion of the treating physician. This adjuvant treatment was applied in 71 cases. Postoperative mortality was defined as death due to any cause within 30 days of the operation. Follow-up of the patients was obtained from our hospital records, or by written questionnaires to the referring pulmonologists or general practitioners.
|
Methods |
|---|
|
TopFootnotesAbstractIntroductionPatients and MethodsMethodsStatistical AnalysisResultsUnivariate Analysis of...Multivariate AnalysisCommentAppendix 1AcknowledgmentsReferences |
|---|
|
Statistical Analysis |
|---|
|
TopFootnotesAbstractIntroductionPatients and MethodsMethodsStatistical AnalysisResultsUnivariate Analysis of...Multivariate AnalysisCommentAppendix 1AcknowledgmentsReferences |
|---|
A Cox regression analysis was carried out with a conditional backward procedure and treating categoric covariates with the deviation method. A significance level of 0.10 was used for adding and deleting. The importance of each selected covariate was evaluated by the Wald statistic, taking into account the remaining selected variables. The relative risk of a covariate (and its 95% confidence interval [CI]) was calculated by the exponent of the regression factor B (and its 95% CI).
In the tables, the median survival time with its 95% CI and the 2- and 5-year survival percentages were represented. We chose these points because most of the deaths occur in the first 2 years after therapy, whereas a survival of 5 years is generally regarded as very suggestive for cure. A p value less than 0.05 was considered significant. To indicate some trends, p values less than 0.20 are mentioned between brackets in the tables.
The statistical program SPSS was used for the analysis of the data.
|
Results |
|---|
|
TopFootnotesAbstractIntroductionPatients and MethodsMethodsStatistical AnalysisResultsUnivariate Analysis of...Multivariate AnalysisCommentAppendix 1AcknowledgmentsReferences |
|---|
The distribution of the metastases in the different MLN is illustrated in Table 1
. The tracheobronchial (level 4), subaortic (level 5), and subcarinal levels (level 7) were the most common levels with metastases. In our 140 patients, 187 metastatic sites were found distributed over the 9 MLN levels, an average of 1.4 per patient.
|
) was 14 months (95% CI, 10 to 18 months). Two-year and 5-year survival rates were 36.6% (95% CI, 28.4% to 44.8%) and 20.8% (95% CI, 17.2% to 24.4%). In a log-rank analysis, the patients treated with postoperative radiotherapy seemed to have a better survival than the others (24.4% at 5 years versus 16.8%; p = 0.02). After correction for the postoperative mortality, this effect was no longer present (p = 0.38).
|
. Right upper or middle lobe tumors most readily spread to the superior mediastinal compartment (levels 1 to 4), right lower lobe tumors to the lower mediastinal compartment (levels 7 to 9), left upper lobe tumors to the aortic compartment (levels 5 and 6), and left lower lobe tumors to all compartments. The spread of a right upper or middle lobe tumor to the lower mediastinal compartment or of a left lung tumor to the aortic as well as the superior mediastinal compartment was an ominous finding: no 5-year survivors were present in these categories.
|
|
Univariate Analysis of Prognostic Factors |
|---|
|
TopFootnotesAbstractIntroductionPatients and MethodsMethodsStatistical AnalysisResultsUnivariate Analysis of...Multivariate AnalysisCommentAppendix 1AcknowledgmentsReferences |
|---|
). This difference was significant in patients staged with mediastinoscopy: 32.2% for mediastinoscopy-negative and 15% for mediastinoscopy-positive ones (p = 0.04).
|
CLINICAL CHARACTERISTICS.
Most of the patients had WHO performance status 1 at diagnosis (Table 4). Patients with WHO performance status 0 had a significantly better 5-year survival (23.8%) than the others (p = 0.02). The difference was significant for WHO 0 versus 1, not for WHO 1 versus 2. Patients without symptoms had a slightly better survival (20.5%) than the symptomatic ones (12.0%). No differences in survival between men and women were found. Weight loss during the 3 months before surgical exploration did not correlate with survival. Even though therapeutic delay (ie, time between the first suspicion on chest roentgenogram and resection) was sometimes more than 12 weeks, no survival effect of the therapeutic delay could be found.
|
Preoperative CEA level, available in 61 patients, correlated with survival: nearly all patients with long-term survival had a CEA value less than 5 ng/mL. If the limit was set at 7.5 ng/mL, only 1 patient had some long-term follow-up and died of tumor after 39 months.
TUMOR CHARACTERISTICS.
Survival was comparable for tumors in the upper, middle or lower lobe (Table 5). Tumors in the main bronchus had a worse 5-year survival (10.6%, nonsignificant difference). Right-sided and left-sided tumors had a nearly identical prognosis. Most patients had central tumors on bronchoscopic examination. Patients with peripheral tumors had a slightly longer median survival time.
|
There were 35 adenocarcinomas, 86 squamous cell, 6 bronchioloalveolar, 7 large cell, and 6 adenosquamous carcinomas. The survival was very similar for all histologic types. It should be noted that the 5-year survival was nil in N2 bronchioloalveolar cell carcinoma. The degree of differentiation of the tumor and the presence of vascular invasion by the tumor did not influence survival.
MEDIASTINAL LYMPH NODE.
Patients with only one metastatic MLN level had a significantly better median survival time (20 months) than those with multiple levels (11 months), but survival curves converged at 5 years (Table 6). The number of metastatic MLN levels was significantly related to survival in patients with nonsquamous tumors (p = 0.03) and patients with a positive mediastinoscopy (p = 0.02). The topography of the metastatic MLN had some importance: patients with metastases in the superior mediastinal compartment only had a better 5-year survival. The differences were not significant, however, in the overall comparison between the five groups.
|
|
Multivariate Analysis |
|---|
|
TopFootnotesAbstractIntroductionPatients and MethodsMethodsStatistical AnalysisResultsUnivariate Analysis of...Multivariate AnalysisCommentAppendix 1AcknowledgmentsReferences |
|---|
Five covariates were significantly related with survival: extent of the primary tumor (relative risk [RR] = 1.43), WHO performance status (RR = 1.37), pathology (RR = 1.29), number of metastatic MLN levels (RR = 1.68), and clinical N status (RR = 1.43). Figure 2 illustrates these relative risks. They are the ratio of the estimated hazard of dying for a patient with an unfavorable characteristic to the hazard for a patient without this characteristic. If the RR is 1, the characteristic does not influence survival; if it is more than 1, the characteristic is associated with an increased hazard of dying, and thus a worse survival. If the 95% CI of the characteristic does not cross the line at value 1, the null hypothesis that this variable is not associated with survival can be rejected, and the finding is representative for the population. For instance, if extent of the tumor is considered, a patient with a pT3/pT4 tumor has a relative risk of 1.43, and thus a 43% greater chance of dying than a patient with a pT1/pT2 tumor. The 95% CI is far from the value-1 line; this finding thus is significant for the entire population.
|
|
Comment |
|---|
|
TopFootnotesAbstractIntroductionPatients and MethodsMethodsStatistical AnalysisResultsUnivariate Analysis of...Multivariate AnalysisCommentAppendix 1AcknowledgmentsReferences |
|---|
The resectability rate (defined as the number of complete resections divided by the total number of resections) was 80.7%. This figure is high and in line with other series using mediastinoscopy for staging [1, 6–8]. The resectability rate was lower but still 63.2% in patients with a positive mediastinoscopy before operation. This can be compared with the 65% resectability rate in the cN2 patients of Pearson and associates [1], but is much higher than the 18% resectability rate reported by Martini and Flehinger [9], who used chest roentgenography and bronchoscopy for mediastinal staging, or the 28% resectability rate mentioned by Watanabe and colleagues [4]. In our view, this reflects our selective attitude in case of a positive mediastinoscopy, but it also confirms that mediastinoscopy is a much better method of staging than CT scanning to avoid futile incomplete resections.
The postoperative mortality of 5.7% is compatible with those of other series. A higher figure was found in the mediastinoscopy-positive patients, where the chances of survival were similar to the risk of postoperative mortality, both in our data and in those from others [1, 4, 9, 10].
The distribution of the metastatic MLN indicated that unforeseen N2 was most frequently found in level 4, level 5 (predominantly in patients with left upper lobe lung tumors staged with cervical mediastinoscopy but not with left anterior mediastinotomy), and level 7. Patterson and associates [8] reported that patients with unforeseen N2 disease in the subaortic lymph nodes only had a 5-year survival of 28% and even 42% if a complete resection was possible. We were not able to confirm these very good results (21% in our series).
As in the anatomic studies by Nohl-Oser, we found that right upper and middle lobe tumors had pathways of lymphatic metastases to the right superior MLN, right lower lobe tumors to the inferior ones, and left upper lobe tumors to the aortic ones. Left lower lobe tumors have a more variable pattern of spread both to the left parasternal MLN and to the subcarinal ones, with a risk of crossing the midline. A lymphatic spread not following the anatomy was a poor prognostic sign: eg, there were no 5-year survivors in patients with a right upper or middle lobe tumor with metastases in the inferior mediastinum.
The long period of follow-up protected against the major pitfall of actuarial survival analysis: making a very early report of the 5-year survival of a group of patients, of whom only a minority have a follow-up approaching 5 years, with a rather unreliable tail of the curve based on only a few patients at risk. In our 140 pN2 patients undergoing resection, the median survival time was 14 months (95% CI, 10 to 18 months) and the 5-year survival was 20.8% (95% CI, 17.2% to 24.4%). This was better than the survival in other published series [1, 4, 9, 11, 12], and can be explained by our very selective attitude toward resection in mediastinoscopy-positive patients. Their proportion was only 19 of 140 cases in this series.
Patients with a preoperative cN0-1 stage had a more favorable survival, 21.6% at 5 years in our group, largely due to the far better survival (32.2%) of patients with a preoperative negative mediastinoscopy. This is in line with the 5-year survival in series of surgically explored patients with a negative preoperative mediastinoscopy: 17.8% reported by Goldstraw and associates [7], 28% by Patterson and associates [8], and 24% by Pearson and associates [1]. This was confirmed in our multivariate analysis.
The importance of a complete resection was logical and in line with literature data [1, 4, 6, 7, 9, 11–13]. Among the 27 patients with incomplete resection, there was only 1 absolute 5-year survivor. He had a negative mediastinoscopy, a T1 tumor, and intracapsular metastatic MLN at two levels. He underwent pneumonectomy plus mediastinal dissection, but with a positive bronchial resection margin.
The WHO performance status proved to be an important prognostic factor: patients with WHO score 0 had a significantly (p = 0.02) better survival. Performance status is a well-known prognostic factor in studies on patients with metastatic NSCLC [14]. In series on surgical treatment, however, it is rarely if ever reported. Nonetheless, we think it is useful in difficult clinical decisions or in the planning of combined modality treatment, given that it was the second most important prognostic factor in our Cox model.
Weight loss influences the prognosis in different cancer types. Weight loss is frequent in patients with NSCLC, and sometimes not due to metastatic disease but to the metabolic derangements of the cancer anorexia and cachexia syndrome, to which NSCLC patients are prone [15, 16]. Surgical series on patients with pN2 disease do not contain sufficient detail on this issue to be compared with our data. We could speculate that the absence of a relationship between weight loss and survival in our pN2 patients is due to the fact that weight loss is caused by metabolic derangements and not by (occult) distant metastatic disease. In this situation, a relationship between weight loss and survival after a complete resection is not necessarily present.
We also examined the effect of therapeutic delay (defined as the interval between the first sign of disease on chest roentgenogram and the surgical exploration). In our pN2 patients we could not confirm the relationship between therapeutic delay and survival as reported in the Ludwig Lung Cancer Study Group data [14]. We were further interested in this effect because the question is sometimes raised if a delay due to waiting times for sophisticated staging or other examinations is not harmful. The fact that our data did not show this can be explained by the long doubling times of the different types of NSCLC. A few weeks are only a fraction in the evolution of this tumor. This is certainly not a plea for loss of time, but against therapeutic rush at the cost of precise medical evaluation and staging of the patient.
Local extent (T stage) was the only tumor characteristic that significantly (p = 0.01) influenced survival in the univariate analysis. In the Cox model, it was the prognostic factor with the highest significance (p = 0.003). The behavior of T1 N2 and T2 N2 disease was similar, as was the case for T3 N2 and T4 N2 disease, where practically no long-term survival was found. This is in agreement with the literature data [1, 4, 8, 9, 11–13, 17, 18]. The report of Goldstraw and colleagues [7] is one where these findings were not present.
Survival was comparable in patients with central versus peripheral tumors at bronchoscopy. Some authors suggested a better survival for peripheral tumors [3, 9], others [19] for central tumors. No firm conclusions can be made on this issue.
Roeslin and associates [20] found vascular invasion by the tumor to be the only prognostic factor in their Cox regression model. In a recent pathologic study, Ichinose and colleagues [21] found an influence of venous invasion in their Cox model. We were not able to confirm this finding.
The pathology of the tumor is reported to be an important factor by some groups [7, 19] but not by others [4, 9, 10, 12, 18, 20, 22]. This can be due to an insufficient number of patients in individual series. When we pooled the literature data on this issue (analysis not yet reported), we found a better survival in patients with squamous cell tumors. Moreover, squamous cell carcinoma had a significantly better prognosis in our Cox model (p = 0.03).
Of the examined MLN characteristics, extracapsular spread was found to be of borderline significance. This is in agreement with some authors [7, 11, 19, 20], but not with others [10, 13, 23]. Some groups [1, 6, 18] do not undertake surgical exploration in patients with extracapsular spread. The presence of capsule rupture was not withheld as a prognostic factor in our Cox model. In an other recent analysis by Ricquet and co-workers [24], capsular rupture had no influence on survival.
The number of metastatic MLN levels did not influence survival in our univariate analysis of the total group, but was important in patients with nonsquamous tumors or with a positive mediastinoscopy. In the Cox model, the number of metastatic MLN levels was a relevant prognostic factor. Literature data also pointed out that this factor is very important for the prognosis [7, 9, 12, 13, 18].
The results on the importance of the distribution of the metastases over the MLN are very conflicting in the literature. One group [25] reported better survival for patients with metastases in the high MLN only; two others [9, 18] for reported better survival cases with low MLN only. In the above-mentioned study by Ricquet and co-workers [24], the location of the metastatic MLN was not important for the prognosis. One notable exception were metastases in the subcarinal region, which led to a significantly worse prognosis in most series [1, 9, 12, 13, 22, 25, 26]. In our patients, the only difference was a better 5-year survival (30.1%) in patients with metastases in the superior mediastinum than in those with subcarinal metastases (18.4%).
At the time of treatment of our patients, there was no unequivocal definition of minimal N2 disease in the literature. Recently, the International Association for the Study of Lung Cancer proposed one positive lower mediastinal nodal station excluding subcarinal adenopathy. We used a somewhat more strict definition of metastasis in only one level, intracapsular, and macroscopically not suspect for metastasis at mediastinoscopy or surgical exploration. Using this definition, we found only a trend toward better survival. However, this distinction proved to be very useful in the group of patients with a positive mediastinoscopy. Each of the few long-term survivors in this category had "strictly minimal N2 disease." Three are still alive, 2 absolute 5-year survivors and 1 patient in follow-up at 36 months. These 3 patients had a T2 right-sided tumor, underwent complete resection, and had MLN metastasis without extracapsular spread in one single level, located in the right paratracheal mediastinum in 2 and subcarinal in 1. If more than minimal N2 disease was found at mediastinoscopy, survival prospects were too poor to justify immediate surgical exploration in these cases.
The prognostic impact of laboratory values such as LDH level is well known in patients with metastatic disease [14]. In earlier disease, however, laboratory values are known to be less important, because disease stage dominates the survival. N2 patients are situated between early (usually surgically treated) and advanced (medically treated) disease; it was thus interesting to look at biological findings. In this group very few, if any, data were found in the classic surgical series. Although we could not find a linear correlation between LDH level and survival, an abnormal (ie, more than 460 U/L in our center) pretreatment value was an ominous finding: all but 1 of these patients died shortly after treatment. A similar influence was found for the preoperative CEA level. Except for 1 individual, who survived 3 years, all patients with a preoperative value of more than 7.5 ng/mL died within 2 years after surgical exploration. The importance of this marker was also noted by others. Concannon and associates [27] found that all of the patients with CEA levels greater than 6 ng/mL died in less than 3 years. Vincent and associates [28] noted that the median survival time for patients with a CEA level less than 2.5 ng/mL was 34 months, whereas it was only 9 months in subjects with a CEA level greater than 5 ng/mL. Finally, Icard and colleagues [29] found no long-term survival in patients with a CEA level greater than 30 ng/mL. Survival at 5 years in 45 patients with stage IIIA disease and a CEA level greater than 10 ng/mL was only 7%.
We conclude that there is a large variation in the literature data on survival and prognostic factors in stage IIIA N2 NSCLC treated by primary surgical resection. With the advantages of a detailed multivariate analysis, we identified the following factors determining the chances for cure in pN2 patients undergoing primary surgical treatment:
Resection was thus rewarding in patients with unforeseen N2 disease, especially after a negative mediastinoscopy. The median survival time of these patients was 20 months, and a 32.2% 5-year survival could be obtained. In this group, survival is expected to be worse in patients with a more advanced T stage, a lower performance status, a nonsquamous histology, involvement of more than 1 MLN level, especially in adenocarcinoma, and a high preoperative value of LDH or CEA. These prognostic factors should be taken into account in adjuvant protocols. In patients with clinical N2 disease discovered at mediastinoscopy, surgical treatment was only worthwhile if there was strictly minimal N2 disease. In other instances a preoperative induction protocol should nowadays be considered.
|
Appendix 1 |
|---|
|
TopFootnotesAbstractIntroductionPatients and MethodsMethodsStatistical AnalysisResultsUnivariate Analysis of...Multivariate AnalysisCommentAppendix 1AcknowledgmentsReferences |
|---|
|
Acknowledgments |
|---|
|
TopFootnotesAbstractIntroductionPatients and MethodsMethodsStatistical AnalysisResultsUnivariate Analysis of...Multivariate AnalysisCommentAppendix 1AcknowledgmentsReferences |
|---|
|
Footnotes |
|---|
|
TopFootnotesAbstractIntroductionPatients and MethodsMethodsStatistical AnalysisResultsUnivariate Analysis of...Multivariate AnalysisCommentAppendix 1AcknowledgmentsReferences |
|---|
Address reprint requests to Dr.Vansteenkiste, Respiratory Tumor Unit, Department of Pulmonology, University Hospital Gasthuisberg, Herestraat 49, B-3000 Leuven, Belgium (e-mail: johan.vansteenkiste{at}uz.kuleuven.ac.be).
|
References |
|---|
|
TopFootnotesAbstractIntroductionPatients and MethodsMethodsStatistical AnalysisResultsUnivariate Analysis of...Multivariate AnalysisCommentAppendix 1AcknowledgmentsReferences |
|---|
Related Article
This article has been cited by other articles:
|
|
 |
|
  P. Misthos, E. Sepsas, J. Kokotsakis, I. Skottis, and A. Lioulias The Significance of One-Station N2 Disease in the Prognosis of Patients With Nonsmall-Cell Lung Cancer Ann. Thorac. Surg., November 1, 2008; 86(5): 1626 - 1630. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Z. Mansour, E. A. Kochetkova, N. Santelmo, X. Ducrocq, E. Quoix, J.-M. Wihlm, and G. Massard Persistent n2 disease after induction therapy does not jeopardize early and medium term outcomes of pneumonectomy. Ann. Thorac. Surg., July 1, 2008; 86(1): 228 - 233. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Dooms, E. Verbeken, S. Stroobants, K. Nackaerts, P. De Leyn, and J. Vansteenkiste Prognostic Stratification of Stage IIIA-N2 Non-Small-Cell Lung Cancer After Induction Chemotherapy: A Model Based on the Combination of Morphometric-Pathologic Response in Mediastinal Nodes and Primary Tumor Response on Serial 18-Fluoro-2-Deoxy-Glucose Positron Emission Tomography J. Clin. Oncol., March 1, 2008; 26(7): 1128 - 1134. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Riquet, P. Bagan, F. Le Pimpec Barthes, E. Banu, F. Scotte, C. Foucault, A. Dujon, and C. Danel Completely Resected Non-Small Cell Lung Cancer: Reconsidering Prognostic Value and Significance of N2 Metastases Ann. Thorac. Surg., December 1, 2007; 84(6): 1818 - 1824. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. Nechala, A. J. Graham, S. D. McFadden, S. C. Grondin, and G. Gelfand Retrospective Analysis of the Clinical Performance of Anterior Mediastinotomy Ann. Thorac. Surg., December 1, 2006; 82(6): 2004 - 2009. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Okamoto, R. Maruyama, R. Suemitsu, Y. Aoki, H. Wataya, M. Kojo, and Y. Ichinose Prognostic value of the histological subtype in completely resected non-small cell lung cancer Interactive CardioVascular and Thoracic Surgery, August 1, 2006; 5(4): 362 - 366. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Mizuguchi, K. Inoue, T. Iwata, T. Nishida, N. Izumi, T. Tsukioka, N. Nishiyama, T. Uenishi, and S. Suehiro High Serum Concentrations of Sialyl Lewisx Predict Multilevel N2 Disease in Non-Small-Cell Lung Cancer Ann. Surg. Oncol., July 1, 2006; 13(7): 1010 - 1018. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Rami-Porta, M. Mateu-Navarro, J. Freixinet, M. de la Torre, A. Jose Torres-Garcia, Y.-W. Pun, A. Canto Armengod, and on behalf of the Bronchogenic Carcinoma Cooperativ Type of resection and prognosis in lung cancer. Experience of a multicentre study Eur. J. Cardiothorac. Surg., October 1, 2005; 28(4): 622 - 628. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Takenaka, S. Ishikawa, K. Yanagihara, R. Miyahara, S. Hasegawa, Y. Otake, Y. Morioka, C. Takahashi, M. Noda, H. Ito, et al. Prognostic Significance of Reversion-Inducing Cysteine-Rich Protein With Kazal Motifs Expression in Resected Pathologic Stage IIIA N2 Non-Small-Cell Lung Cancer Ann. Surg. Oncol., October 1, 2005; 12(10): 817 - 824. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Gawrychowski, K. Brulinski, E. Malinowski, and B. Papla Prognosis and survival after radical resection of primary adenosquamous lung carcinoma Eur. J. Cardiothorac. Surg., April 1, 2005; 27(4): 686 - 692. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
