|
|
||||||||
a Department of Thoracic Surgery, Sun Yat-sen University Cancer Center, Guangzhou, Guangdong Province, People's Republic of China
b State Key Laboratory for Cancer Research in South China, Guangzhou, Guangdong Province, People's Republic of China
c Shanghai Lung Tumor Clinical Medical Center, Chest Hospital Affiliated to Shanghai Jiao Tong University, Shanghai, People's Republic of China
Accepted for publication April 8, 2009.
* Address correspondence to Dr Jian-hua Fu, Department of Thoracic Surgery, Sun Yat-sen University Cancer Center, No. 651, Dongfeng East Rd, Guangzhou City, Guangdong Province, 510060, People's Republic of China (Email: j_hfu{at}yahoo.com.cn).
| Abstract |
|---|
|
|
|---|
Methods: Consecutive patients with T4 NSCLC with mediastinal involvement who underwent thoracotomy (n = 146) were retrospectively analyzed. Actuarial survival was calculated. Factors associated with overall survival were identified.
Results: Four operative deaths occurred in pneumonectomy patients. The overall 5-year survival rate was 22.7%; median survival was 24.8 months. Factors favoring survival included complete resection (p < 0.001), N2-negative disease (p < 0.001), and pulmonary great vessel invasion (p < 0.001). Multivariate analysis of 131 patients undergoing lung resection found three factors associated with increased survival: pulmonary great vessel involvement (relative risk [RR] of death, 0.365; 95% confidence interval [CI], 0.208 to 0.639; p < 0.001), R0 resection (RR, 0.387; 95% CI, 0.209 to 0.714; p = 0.002), and postoperative chemotherapy (RR, 0.255; 95% CI, 0.134 to 0.487; p < 0.001). Male sex (RR, 2.962; 95% CI, 1.479 to 5.934; p = 0.002) and N2-positive disease (RR, 3.220; 95% CI, 1.640 to 6.323; p < 0.001) were associated with decreased survival.
Conclusions: N2-positive, T4 NSCLC with mediastinal involvement is not suitable for resection. T4 NSCLC patients with pulmonary great vessel involvement had better survival rates than other T4 subgroups. Pneumonectomy should be done with caution due to a high mortality risk and poor prognosis. Further studies are warranted to assess the role of sex on survival.
| Introduction |
|---|
|
|
|---|
Previous studies suggested that for selected T4 non-small cell lung cancer (NSCLC) with mediastinal involvement, resection might provide a survival benefit [7–13]. However, because of different inclusion criteria and different case numbers, the risk factors and prognosis of surgically treated T4 NSCLC patients with mediastinal involvement remain poorly characterized. For example, considerable differences of opinion exist about the role of pneumonectomy, lymph node status, different pathologic types of the tumor, and various T4 types on survival of T4 NSCLC patients [7–13]. Therefore, patient selection is very important whenever surgical resection is to be performed in patients with T4 NSCLC.
We conducted this study to assess the survival of patients with surgically treated T4 NSCLC with mediastinal involvement to clarify the clinicopathologic features and prognostic factors in this patient population and to propose subgroups of patients with T4 NSCLC with mediastinal involvement who would be more likely to benefit from surgical resection.
| Patients and Methods |
|---|
|
|
|---|
Patient Selection
The study was conducted using data from a prospectively collected NSCLC database, which is composed of 2773 consecutive cases of thoracotomy performed at the Thoracic Surgery Department of Sun Yat-sen University Cancer Center between January 1997 and January 2008. In this electronic database, we found 216 patients (7.8%) diagnosed with pathologically proven, T4 NSCLC according to the 1997 T N M classification system, and 146 of these had T4 disease with mediastinal involvement. We conducted a retrospective survival analysis for the 146 patients and in the patients who underwent surgical resection.
The study excluded patients with bronchioloalveolar carcinoma, small cell lung cancer, T4 NSCLC with malignant effusion, and those with a history of concurrent malignant disease or other previous primary cancer. Patients with a new-onset hoarseness indicating recurrent laryngeal nerve involvement are not surgical candidates in our institute. We excluded patients with intralobar metastases due to the reclassification of these patients as T3 instead of T4 in the coming 2009 T N M classification [14]. Patients with synchronous, multiple primary lung cancers were also excluded according to the criteria proposed by Martini and Melamed [15]. The histologic diagnosis of tumors was based on the World Health Organization criteria [16].
All patients underwent preoperative evaluation that included a chest computed tomography (CT), brain CT, or magnetic resonance imaging (MRI), abdominal ultrasonography or CT scan, and bone scan. Mediastinal lymph node involvement (short diameter larger than 1 cm) was excluded preoperatively based on the CT scan. Mediastinoscopy and positron emission tomography (PET) were not routinely performed as part of the preoperative workup. Mediastinoscopy showed only 6 patients had negative mediastinal lymph nodes, and only 24 patients received PET scans. Routine biochemical profile, fibrobronchoscopy, pulmonary function tests with spirometry, and arterial blood gas analysis at rest were required in all patients. All tumors were thought to be resectable based on discussion by a panel of surgeons before the operation.
Patient Demographics
Most patients were men. The clinical stage before resection was 3 with C T2 N0 M0, 5 with C T2 N1 M0, 22 with C T3 N0 M0, 5 with C T3 N1 M0, 87 with C T4 N0 M0, and 24 with C T4 N1 M0, 24. Table 1
gives the clinicopathologic characteristics of the patients. The primary tumor was located at the right upper lobe in 62, the right middle lobe in 5, the right lower lobe in 26, the left upper lobe in 28, and the left lower lobe in 25. The sites of mediastinal involvement were pulmonary great vessels in 82, trachea in 21, carina in 8, superior vena cava (SVC) in 11, left atrium in 15, aorta in 4, esophagus in 3, and vertebral body in 2.
|
Surgical Procedure
Among the 146 patients with T4 disease with mediastinal invasion who underwent thoracotomy, 131 underwent lung resection and 15 underwent only exploratory thoracotomy because the disease was identified as unresectable. The posterolateral thoracotomy was routinely used, and the type of lung resection is summarized in Table 2. No extracorporeal circulation was used.
|
Systematic mediastinal lymph node dissection or sampling was performed during the operation. For patients undergoing exploratory thoracotomy without lung resection, mediastinal sampling was performed to achieve an accurate pathologic stage. The residual tumor was defined as R0 (no residual tumor), R1 (microscopic residual tumor), or R2 (gross residual tumor). We applied the definition of complete surgical resection proposed by the International Association for the Study of Lung Cancer Staging Committee [17].
Patient Follow-Up
Operative death was defined as death within 30 days of the operation or any time after the operation if the patient did not leave the hospital alive. A follow-up examination was generally performed every 2 months during the first year, every 3 months during the second year, and every 4 months thereafter. If the patient had specific symptoms, the examination was performed as soon as possible for a more careful assessment. The follow-up examination included a physical examination, blood chemistry, and chest roentgenogram. All patients routinely received a chest CT scan, head CT scan or MRI, and radionuclide bone scan.
Local recurrence was defined as tumor relapse in the intrathoracic lymph nodes or surgical residue after R0 resection. Any other site of tumor relapse was defined as distant organ metastasis. The median time from operation to the last contact for the entire cohort was 62.3 months (range, 0.7 to 101 months).
Statistical Analysis
Statistical analysis was performed using SPSS 16.0 software (SPSS Inc, Chicago, IL). Survival was calculated by the Kaplan-Meier method, and the log-rank test was used to assess differences in survival between groups. A two-sided value of p < 0.05 was considered statistically significant. Patients alive at the end of the study were censored for purpose of data analysis. The relationships between the survival and clinicopathologic variables were determined by Cox's multivariate analyses.
| Results |
|---|
|
|
|---|
|
Disease Relapse and Overall Survival
Among 106 patients with R0 resection, 61 had disease relapse during the last follow-up contact. This included metastases to the brain in 16, lung in 14, bone in 9, liver in 6, multiple organs in 10, and 6 patients with local recurrence. Survival data was collected on each patient from the date of operation. The overall survival rates for the entire cohort (including the operative deaths) at 2, 3, and 5 years were 51.2%, 33.7%, and 22.7%, respectively, with a median survival of 24.8 months (Fig 1).
|
|
|
|
The multivariate Cox regression analysis for the 146 patients demonstrated that pulmonary great vessel invasion, N2 negative status, and R0 resection were independent factors for increased overall survival (Table 4). When only the 131 patients with lung resection were involved in multivariate Cox regression analysis, however, the independent factors that favored the overall survival were female sex, pulmonary great vessel invasion, R0 resection, N2 negative status, and postoperative chemotherapy (Table 5).
|
|
| Comment |
|---|
|
|
|---|
The surgical treatment of T4 NSCLC necessitates a more extended resection relative to that of early-stage NSCLC patients, contributing to a high postoperative morbidity and mortality. Previous data showed that the morbidity rate varied from 20% to 50% at different institutions [7–13]; this variability may be due to different definitions of morbidity and different case numbers. In our data, the morbidity rate was 53.4%, which was comparable with previous reports [7–13]. The most common complications were pulmonary and cardiovascular disorders such as pneumonia, atelectasis, persistent air leak, bronchopleural fistula, and heart arrhythmia.
A review of radical resections for T4 lung cancer showed increased postoperative mortality as evidenced by mortality rates of 8% at 30 days and 18% at 90 days [18]. Pneumonectomy contributed to all of the operative deaths in our patients, suggesting that more careful patient selection and more intensive postoperative care should be given to these patients.
Early data showed that extended resections such as carinal resection, sleeve pneumonectomy, and great vessel resection carried a operative mortality of no less than 10%, even in experienced centers [19, 20]. Nevertheless, more recent reports have shown that the operative mortality rates have decreased dramatically, to less than 8% [8, 11, 21].
We did careful preoperative physiologic assessment of the patients, including complete pulmonary function tests and quantitative ventilation-perfusion lung scans. For patients who were potential candidates for pneumonectomy, a bilateral pulmonary radioisotope scan was performed to estimate the remaining lung volume after pneumonectomy. We believe the decreased operative mortality was due to careful patient selection, improvement of surgical experience, and perioperative care.
Survival at 5 years greatly varies among institutions. Bryant and colleagues [22] conducted a study of patients with pathologically proven, node-negative T4 lesions. The 25 patients with mediastinal invasion had a 5-year survival rate of 30%. A similar result was also observed in the study reported by Pitz and colleagues [12]. In our study, the overall survival for the entire cohort was 22.7%, which was comparable with that reported by Osaki and colleagues [11]. The different results reported in different studies may be due to selection bias. When only completely resected (R0) patients were analyzed, the overall survival was as high as 40%; however, no patients with incomplete resection survived to 5 years.
We failed to find a head-to-head, prospective, randomized clinical trial to compare the survival outcome of surgical intervention with that of chemoradiotherapy for operable T4 N0-1 NSCLC. Nevertheless, Albain and colleagues [23] reported that the 5-year survival was only 15% for pathologic T4 N0-1 M0 stage IIIB NSCLC who received chemoradiotherapy in subgroup analysis of a Southwest Oncology Group phase II 9019 study. This suggests that surgical intervention should be considered first for operable T4 N0-1 NSCLC, provided that the patients can medically tolerate the procedure.
Previous reports from trials of induction therapy have suggested that pneumonectomy is a negative risk factor for postoperative survival [24–26]. Martin and colleagues [24, 25] reported the long-term results of combined-modality therapy in resectable NSCLC in 470 patients who underwent thoracotomy after induction chemotherapy with or without induction radiotherapy. The total mortality was 7 of 297 (2.4%) lobectomies and 11 of 97 (11.3%) pneumonectomies in their study. Using multivariate analysis, they determined that pneumonectomy adversely influenced survival [24, 25].
Deslauriers and colleagues [27] conducted a large case number analysis of survival after sleeve lobectomy vs pneumonectomy for different stages of lung cancer with 1230 patients involved, including 1046 pneumonectomies. For the entire group, survival at 5 years was 52% after sleeve lobectomy and 31% after pneumonectomy (p < 0.0001) [27]. In our series, the 5-year survival was also lower for patients in the R0 resection group who underwent pneumonectomy than patients who did not. Although the survival difference was not statistically significant, the borderline significant p = 0.059 may be due to the small number in each group. Patients undergoing pneumonectomy usually had more advanced disease than those who underwent lobectomy or bilobectomy (right lung). In addition, pneumonectomy meant much more loss of pulmonary reserve, possibly contributing to the poor survival. We believe that pneumonectomy should be done with caution for T4 NSCLC because of the high mortality risk and poor prognosis.
Nodal status is generally considered to be an important prognostic factor in NSCLC. Some data have shown that N2-positive T4 NSCLC has a poor prognosis [11, 28, 29]. In contrast, other data showed that N2 lymph node involvement was not a prognostic factor in T4 patients because T4 was a more important determinant of survival than N2 [8, 12, 30]. The median survival time for N2-positive patients in our series was less than 2 years, whereas the long-term survival for pathologic N0 and N1 disease was much better than for N2-positive patients. Multivariate analysis confirmed that N2 was an independent negative risk factor for survival, suggesting that an operation may not be a reasonable choice for N2-positive T4 NSCLC. Thus, the key point is how to rule out N2-positive patients from an operation.
The small number of N0 patients in our study must be considered when interpreting our data. Mediastinoscopy was not a routine part of the preoperative workup. A negative mediastinal lymph node on CT scan was defined as having a short diameter of less than 1 cm. However, most patients were N2-positive according to the postoperative pathologic examination. One possible explanation is that most of the tumors were centralized lesions, which may fuse the enlarged mediastinal lymph node into one mass and make it difficult to discriminate the enlarged lymph node from the primary lesion.
In industrialized countries, PET and mediastinoscopy examinations are routine in most NSCLC patients for a more accurate preoperative staging. We believe these examinations are helpful. However, in developing countries with large populations, such as in China, PET and mediastinoscopic examinations have not been widely used because of the expense to patients, and the PET examination has not yet been covered by medical insurance. We believe that with economic growth, PET examination will be widely used for lung cancer diagnosis and staging in developing countries. A systematic review of cervical mediastinoscopy involving 6505 lung cancer cases showed that the sensitivity was 78% and specificity was 100% [31]. We should, therefore, recommend mediastinoscopic examination to perform an accurate mediastinal lymph node assessment for potentially operable T4 NSCLC.
We found that the site mediastinal involvement in T4 NSCLC was the pulmonary great vessels in 82 of 146 patients (56.2%). When the pericardium was opened, many of these tumors could be completely resected and good patient survival was achieved. Multivariate Cox regression analysis showed that patients with pulmonary great vessel involvement had better survival rates than patients with other organ invasion. This indicates that tumor may be more extended and malignant if it invades the structures beyond the pulmonary great vessels.
Meta-analysis showed that neoadjuvant chemotherapy might give survival benefit for some NSCLC patients [32, 33]; however, the role of neoadjuvant chemotherapy on T4 NSCLC is still unclear. Only 12 patients received neoadjuvant chemotherapy in our series, and the number was not large enough to analyze the effectiveness of neoadjuvant chemotherapy.
Men had a statistically significant decreased survival relative to women in multivariate Cox regression analysis in patients undergoing lung resection. However, the survival difference was not statistically significant in Kaplan-Meier analysis. This effect may be due to the small number of women. We do not know why there were so many more men than women in our series. Further research on sex and survival are warranted.
We concluded that N2-positive patients are not suitable to surgical intervention for T4 NSCLC. T4 patients with pulmonary great vessel involvement had improved survival relative to patients with other mediastinal organ invasion. In addition, pneumonectomy should be done with caution in T4 NSCLC due to a high mortality risk and poor prognosis. Women seem to have better survival rate than men in operable T4 NSCLC; however, because of the small number of women, further research is warranted.
| Acknowledgments |
|---|
|
|
|---|
| Footnotes |
|---|
|
|
|---|
| References |
|---|
|
|
|---|
Related Article
This article has been cited by other articles:
![]() |
F. Stella, A. Dell'Amore, G. Caroli, G. Dolci, N. Cassanelli, G. Luciano, F. Davoli, and A. Bini Surgical results and long-term follow-up of T4-non-small cell lung cancer invading the left atrium or the intrapericardial base of the pulmonary veins Interact CardioVasc Thorac Surg, January 22, 2012; (2012) ivr160v1. [Abstract] [Full Text] [PDF] |
||||
![]() |
H. Igai and H. Yokomise Intrapericardial approach via median sternotomy for hilar non-small cell lung cancer invading the main pulmonary artery Interact CardioVasc Thorac Surg, October 1, 2010; 11(4): 385 - 386. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. Tanju and S. Dilege Extended resections for primary lung cancer with oncological principles Interact CardioVasc Thorac Surg, October 1, 2010; 11(4): 387 - 387. [Full Text] [PDF] |
||||
![]() |
A. Chambers, T. Routledge, A. Bille, and M. Scarci Does surgery have a role in T4N0 and T4N1 lung cancer? Interact CardioVasc Thorac Surg, October 1, 2010; 11(4): 473 - 479. [Abstract] [Full Text] [PDF] |
||||
![]() |
T. G. Lesser Endothoracic Sonography Improves the Estimation of Operability in Locally Advanced Lung Cancer Ann. Thorac. Surg., July 1, 2010; 90(1): 217 - 221. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. D. Sihoe Invited Commentary Ann. Thorac. Surg., August 1, 2009; 88(2): 378 - 379. [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |