|
|
||||||||
Ann Thorac Surg 2007;84:952-958
© 2007 The Society of Thoracic Surgeons
a Thoracic Surgery Service, Hospital General Universitario Gregorio Marañón, Madrid, Spain
b Anatomical Pathology Service, Hospital General Universitario Gregorio Marañón, Madrid, Spain
Accepted for publication April 23, 2007.
* Address correspondence to Dr Simón, Cirugía Torácica, Hospital Gregorio Marañón, C/ Dr. Esquerdo 46, Madrid, 28007, Spain (Email: cm_simon{at}telefonica.net).
| Abstract |
|---|
|
|
|---|
Methods: A retrospective review of a prospective multi-institutional database of patients operated on for lung cancer was undertaken. In all, 1,475 patients with pathologic stage I or II NSCLC were studied (421 underwent pneumonectomy; 1,054 had a lobectomy/bilobectomy). Survival and impact of side of surgery for pneumonectomy and lesser resection groups were analyzed and compared using the Kaplan-Meier method and the Cox proportional hazards model.
Results: Median survival was worse after pneumonectomy than after less extensive resections for patients overall (33 versus 57 months) and for those with stage I NSCLC (38 versus 70 months); however, median survival was better after pneumonectomy for stage II left tumors (55 versus 19 months). Pneumonectomy was an independent adverse determinant of survival for both stage I right tumors (p < 0.001) and stage I left tumors (p < 0.001), but was associated with improved survival for stage II left tumors (p = 0.009).
Conclusions: Pneumonectomy was found to be an independent determinant of survival in patients with stage I and II NSCLC, but results differed for right- and left-sided tumors. Further studies of survival comparing pneumonectomy with lesser resections should differentiate between right and left procedures.
| Introduction |
|---|
|
|
|---|
On the other hand, it is well known that right lung contribution to overall lung function is normally higher than that of the left lung. Moreover, several studies have found that patients who underwent a right-sided pneumonectomy, rather than a left-sided pneumonectomy, have been shown to have higher morbidity and mortality [8–10]. While it is reasonable to anticipate differences in long-term survival after right and left pneumonectomy, little has been reported in the medical literature [11].
The purpose of this study was to reassess the impact of pneumonectomy on long-term survival in patients with pathologic stage I and II NSCLC, and to study the side of pneumonectomy as a possible confounding variable when trying to identify independent prognostic factors for survival across resected stage I and II NSCLC.
| Material and Methods |
|---|
|
|
|---|
General Methodology
From October 1993 to September 1997, every patient with bronchogenic carcinoma who underwent a thoracotomy in a GCCB-S hospital was registered prospectively [12]. The annual cumulative number of cases was close to 50% of the patients operated on for lung cancer in Spain during the same time period. The GCCB-S centers were representative of hospitals throughout Spain providing a wide variety of services and activities [13]. The final registry included 2,994 patients.
All GCCB-S hospitals used similar criteria to assess the functional operability of patients and the oncological operability of the tumors [14]. Patients were staged according to the 1997 TNM staging classification system [15]. Pathologic N0 was classified by mediastinal nodal dissection or sampling of at least four lymph node areas (2 [only in right lung tumors]; 4, 7, and 10 on the same side as the tumor), especially in pT3 [16]. Operative mortality included death within or outside the hospital within 30 days of operation, or death at anytime after operation if the patient did not leave the hospital. All long-term survivors were followed up for more than 5 years. Internal and external audits were conducted [17].
Specific Methodology
After pulmonary resection with curative intent, 1,642 patients were pathologically diagnosed as stage I or II NSCLC. A total of 421 patients (25.6%) underwent pneumonectomy, 1,054 (64.2%) had either a lobectomy or bilobectomy, and 167 (10.2%) had a less extensive resection. Patients with resections less extensive than lobectomy were excluded from the study. Consequently, the study included a total of 1,475 patients (1,375 men), with a median age of 66 years (range, 36 to 87). A total of 1,101 cases was classified stage I (IA 221, IB 880) and 374, stage II (IIA 31, IIB 343). Histologic types included 922 (62.5%) squamous cell carcinomas, 357 (24.2%) adenocarcinomas, 90 (6.1%) large cell carcinomas, 58 (3.9%) unspecified non-small cell carcinomas, and 48 (3.3%) bronchoalveolar carcinomas. According to the standard of care during the time period of the study, no patient received induction treatment, and 216 patients (14,6%) received some kind of adjuvant therapy.
Definitions and Statistics
Eleven variables were tested with univariate and multivariate analysis with the endpoint being survival: age, sex, pathologic TNM stage (pTNM), squamous cell carcinoma, comorbidity, smoking, preoperative plasma hemoglobin, preoperative forced expiratory volume in one second (FEV1), side of surgery, tumor size, and type of resection (pneumonectomy versus lesser resection). Variables were selected for analysis based on an extensive literature review, clinical experience, and availability and completeness of data in the registry (missing data for each variable <5%).
The presence of comorbidity was considered when chronic obstructive pulmonary disease, arterial hypertension, previous tumor, cardiac disease, peripheral vascular disease, or diabetes appeared alone or as an associated disease. Chronic obstructive pulmonary disease was defined when a compatible clinical picture existed, when there was a chronic obstruction to the air flow, and when other conditions with similar symptoms were ruled out. Obstruction to the air flow was confirmed when the postbronchodilator FEV1/FVC was less than 0.7 [18]. Any type of neoplasia occurring in any previous period of time was registered as a previous tumor, excluding nonmelanoma skin tumors. The presence of cardiac disease was considered when there was valve disease, heart failure, ischemic heart disease, or a combination of these.
Continuous data are presented as mean with their standard deviation, and categorical variables as percentages. To compare the frequency distribution of the tested variables across different populations, either the Pearson
2 technique or the Fishers exact test was used for qualitative variables, and the Student t test and the Levene test when required were used for quantitative variables. A p value of less than 0.05 was considered significant.
For the prognostic analysis, survival was calculated from the date of operation for each type of resection (pneumonectomy versus lesser resection) using the Kaplan-Meier method. Survival also was calculated for each type of resection (right or left pneumonectomy versus lesser resection) and for each stage, considering the side of resection. Survival rates are presented with their standard error. Differences in survival between groups were tested for significance using the log-rank test. The univariate Cox regression model was used to assess the impact of selected factors on long-term survival. The hazard ratio (HR) of a univariate variable (and its 95% confidence interval [CI]) was calculated by the exponent of the regression factor B (and its 95% CI), and every variable showing marginal significance (p < 0.2) was then tested in a Cox multivariate model. A Cox proportional hazards model was fit to examine and adjust for the effects of surgical treatment and other covariates on survival. The backward stepwise procedure and the likelihood ratio test were used to select the variables with the greatest prognostic value (p < 0.05). The hazard ratio of a covariate (and its 95% CI) was calculated by the exponent of the regression factor B (and its 95% CI). A p value of less than 0.05 was considered significant. The statistical package SPSS PC, version 13.0 (SPSS, Chicago, Illinois) was used for analysis.
| Results |
|---|
|
|
|---|
|
|
|
|
|
| Comment |
|---|
|
|
|---|
Appropriate surgical management for centrally or locally advanced tumors is debatable. Pneumonectomy may be inevitable owing to anatomical and technical considerations, but it is associated with significant morbidity and mortality [1, 2]. This outcome was observed in our series, where mortality after pneumonectomy exceeded 13%.
Reports from the Mayo Clinic [5] and St. Marguerite Hospital [6] implicated pneumonectomy as an independent adverse factor for survival in early disease stages of lung cancer. Other series could not identify pneumonectomy as a predictor of survival in stage I [20, 21] or in stage T2N1M0 [22]. Luzzi and coworkers [23] found that in stage II NSCLC (T1-2N1) with station 10 involved, pneumonectomy showed a better 5-year survival (58%) than lobectomy (33%), mostly because of a lower incidence of local recurrence in the pneumonectomy group (0% versus 24%).
Ferguson and Karrison [2] reviewed the survival of NSCLC patients who underwent lobectomy/bilobectomy (n = 340) or pneumonectomy (n = 102). Overall (stage I to III), median and 5-year survivals for lobectomy/bilobectomy were significantly greater than those for pneumonectomy (38 versus 17 months and 43% versus 30%, respectively). After adjusting for age, primary tumor status, regional nodal status and preoperative FEV1, survival did not differ significantly between the types of operations.
In another study, 485 patients with stage I NSCLC underwent either pneumonectomy (n = 111) or a smaller resection than pneumonectomy (n = 374). The 5-year survival rate after pneumonectomy was 44%, while after less extensive resection survival was 50%, with a significant survival benefit in the lesser resection group. On multivariate analysis, pneumonectomy, T2 stage, older age, and male sex were independent adverse predictors of survival [7].
Parenchyma-sparing techniques, such as sleeve lobectomy, have evolved as an alternative to pneumonectomy in selected patients with central tumors [24]. A meta-analysis of results of sleeve lobectomy and pneumonectomy in stage I and II NSCLC suggested that performance of a sleeve lobectomy rather than a pneumonectomy for patients who are anatomically appropriate, regardless of underlying cardiopulmonary status, provided an overall survival and quality of life advantage [25]. In a series of 1,230 patients who underwent either pneumonectomy (n = 1,046) or sleeve lobectomy (n = 184) for lung cancer stage I to III, univariate analysis showed that 5-year survival was higher after sleeve lobectomy for the entire group (31% versus 52%) and across stage I (50% versus 66%) and II (34% versus 50%) [26]. Likewise, other authors have shown that survival after sleeve resection appears to be no different or is even better than survival after pneumonectomy [27, 28].
In our series, similar to previous reports, univariate and multivariate analyses identified pneumonectomy as an independent adverse factor for survival in stage I NSCLC; yet, after adjusting for age, tumor size, comorbidity, and type/side of surgery interaction, pneumonectomy was found to improve long-term survival in stage II patients. This finding differs from the aforementioned results that favored sleeve lobectomy for stage II NSCLC patients [26]. However, our findings may not be comparable with these results because sleeve lobectomies were rarely performed by GCCB-S surgeons before 1997.
The influence of side of surgery when comparing pneumonectomy to lesser resection for lung cancer has seldom been studied. Liu and colleagues [11] reported a better 5- and 10-year survival after left pneumonectomy (36.7% and 9.5%, respectively) than after right pneumonectomy (21.7% and 8.3%) for lung cancer. Deslauriers and coworkers [26] reported no survival differences among patients with resected stage I to III NSCLC, when comparing left and right pneumonectomy (30% versus 32% 5-year survival, respectively). In the present series, multivariate analysis in the overall study population and in patients with stage I and II suggested a significant interaction between type and side of surgery. To further clarify this interaction, we stratified patients into right and left side surgery groups. In patients with resected stage I NSCLC, both right and left pneumonectomy were independent adverse predictors of survival when compared with lesser resections. However, on univariate and multivariate analysis, left pneumonectomy was significantly associated with improved survival after resection of stage II NSCLC when compared with lesser left resections.
These paradoxic results have not been reported previously and may be explained by two opposed effects of pneumonectomy: the long-term consequences of pneumonectomy (especially right pneumonectomy) on cardiorespiratory function and, as suggested by Luzzi and colleagues [23], the reduced risk of local relapse, possibly due to a better clearance of N1 lymph nodes. If this were so, pneumonectomy would not provide any oncologic advantages for stage I patients, and it also would reduce long-term survival. In stage II patients, the considerable repercussions of right pneumonectomy would minimize the beneficial effect of a larger oncologic radicality. However, with left pneumonectomy, which is better tolerated, an evident survival benefit would be provided by reducing tumor recurrence. In fact, in our series 5-year survival in patients with stage II disease after left pneumonectomy (47.7%) was similar to survival in stage I patients (45.5%), and even higher than 5-year survival in stage I patients after right pneumonectomy (32.7%). Thus, it appears that the effect of pneumonectomy on survival is an even more powerful factor than stage when disease is localized to the inside of the lung.
Information regarding precise cause of late death may have been useful to rule out the contribution of pneumonectomy on noncancer-related deaths. Unfortunately, in the present series accurate data about cause of death was not available for many cases. Another limitation of the study is that with our series of cases we do not know if these results would be reproducible after broncoplastic resections. It is possible that a more complete resection of N1 lymph nodes and a wider resection margin might contribute to good published results with sleeve lobectomies as opposed to pneumonectomy.
Our findings suggest that pneumonectomy is an independent adverse determinant of survival in patients with stage I NSCLC, but that left pneumonectomy seems to improve the long-term survival in patients with stage II disease. We believe that further prospective studies of survival comparing pneumonectomy with lesser resections, such as bronchoplastic procedures, should differentiate between right and left procedures.
| Appendix |
|---|
Coordinators (hospital): José-L Duque, MD (Universitario, Valladolid); Angel López-Encuentra, MD (12 de Octubre, Madrid); and Ramón Rami, MD (Mutua Terrassa, Barcelona). Local representatives (hospital): Julio Astudillo, MD; Pedro López de Castro, MD (Germans Trias i Pujol, Barcelona); Josep-María Gimferrer, MD (Clinic, Barcelona); Antonio Cantó, MD; Antonio Arnau, MD (Clínico, Valencia); Juan Casanova, MD; Manuel Mariñán, MD (Cruces, Bilbao); Manuel Castanedo, MD (Universitario, Valladolid); Antonio Fernández de Rota, MD (Carlos Haya, Málaga); Federico González-Aragoneses, MD; Nicolas Moreno, MD (Gregorio Marañón, Madrid); Jorge Freixinet, MD (Doctor Negrín, Las Palmas); Nicolás Llobregat, MD (del Aire, Madrid); Nuria Mañes, MD (Jiménez Díaz, Madrid); Mireia Serra, MD; José Belda, MD (Mutua de Terrassa, Barcelona); José-L Martín de Nicolás, MD (12 de Octubre, Madrid); Nuria Novoa (Complejo Hospitalario, Salamanca); Jesús Rodríguez, MD (Complejo Hospitalario, Oviedo); Antonio-José Torres-García, MD; Ana Gómez, MD (San Carlos, Madrid); Mercedes de la Torre (Juan Canalejo, La Coruña); Abel Sánchez-Palencia, MD; Javier Ruíz-Zafra, MD (Virgen de las Nieves, Granada); Andrés Varela-Ugarte, MD; Francisco-J Moradiellos, MD (Puerta de Hierro, Madrid); and Yat Wah Pun, MD (La Princesa, Madrid). Data analysis: Agustín Gómez de la Cámara, MD, and Francisco Pozo-Rodriguez, MD.
| Acknowledgments |
|---|
|
|
|---|
| Footnotes |
|---|
|
|
|---|
| References |
|---|
|
|
|---|
Related Article
This article has been cited by other articles:
![]() |
E. J. Veen, M. L.G. Janssen-Heijnen, E. D. Ritchie, B. Biesma, M. P.H. van den Bogart, and R. Jan Bolhuis Pneumonectomy for bronchogenic carcinoma: analysis of factors predicting short- and long-term outcome Interactive CardioVascular and Thoracic Surgery, August 1, 2009; 9(2): 260 - 264. [Abstract] [Full Text] [PDF] |
||||
![]() |
R. J. Cerfolio Invited commentary Ann. Thorac. Surg., September 1, 2007; 84(3): 958 - 958. [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 |