EJCTS Click here for details of sales representative
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to Personal Folders
Right arrow Download to citation manager
Right arrow Permission Requests
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Ma, Z.
Right arrow Articles by Cheng, H.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Ma, Z.
Right arrow Articles by Cheng, H.
Related Collections
Right arrow Lung - cancer
Right arrow Trachea and bronchi

Eur J Cardiothorac Surg 2007;32:20-28. doi:10.1016/j.ejcts.2007.03.018
Copyright © 2007, European Association for Cardio-Thoracic Surgery. Published by Elsevier B.V. All rights reserved

Reviews

Does sleeve lobectomy concomitant with or without pulmonary artery reconstruction (double sleeve) have favorable results for non-small cell lung cancer compared with pneumonectomy? A meta-analysis

Zhiyuan Maa,b, Aiqiang Donga,*, Junqiang Fana, Haifeng Chenga

a Department of Cardiothoracic Surgery, Second Affiliated Hospital, School of Medicine, Zhejiang University, Hangzhou 310009, China
b Department of Thoracic and Cardiovascular Surgery, Shanghai Jiao Tong University Affiliated First People's Hospital, Shanghai 200080, China

Received 1 January 2007; received in revised form 7 March 2007; accepted 13 March 2007.

* Corresponding author. Tel.: +86 571 87783641; fax: +86 571 87022660. (Email: dr_dongaiqiang{at}sina.com).


   Abstract
Top
 Abstract
1. Introduction
 2. Materials and methods
 3. Results
 4. Discussion
 References
 
It has been reported that sleeve lobectomy (SL) concomitant with or without pulmonary artery reconstruction (PAR) might be an alternative procedure for pneumonectomy (PN) in non-small cell lung cancer (NSCLC). The aim of this study was to assess whether SL or PN offers a low morbidity and mortality and better long-term survival. We performed a meta-analysis of studies published in English between 1996 and 2006 to comprehensively compare the postoperative mortality, morbidity, locoregional recurrences, and time-to-event outcomes of SL and PN in NSCLC, and reviewed the recent literatures on PAR in the corresponding period as well. Twelve studies met the defined criteria including a total of 2984 subjects, and five studies for PAR. The odds ratio for postoperative mortality (SL vs PN) was 0.65 (95% confidence interval (CI): 0.42–1.01), 1.01 (95% CI: 0.70–1.44) for postoperative complications, and 0.91 (95% CI: 0.45–1.82) for locoregional recurrences. The weighted mean operative mortality for PAR was 3.3%, and 32.4% for complications. The estimated combined hazard ratio for overall survival in 10 studies was 0.70 (95% CI: 0.62–0.79) in favor of SL group. The median overall survival was 60 months for the SL group, 26 months for the PN group, and 30 months for PAR group. Survival difference in patients with pN0 or pN1 at 1 year demonstrated a pooled risk difference (SL vs PN) of 0.03 (95% CI: –0.08–0.13), 0.13 (95% CI: 0.00–0.25) in patients with pN2 at 1 year, 0.21 (95% CI: 0.07–0.36) in patients with pN0 or pN1 at 5 years, and 0.06 (95% CI: –0.10–0.21) in patients with pN2 at 5 years. Our results suggests that SL with or without PAR can be accomplished safely in selected patients without increasing the morbidity and mortality as compared to PN, that SL even with PAR offers better long-term survival than does PN, and that a more radical operation such as PN is not a more appropriate procedure, even in higher stage tumors.

Key Words: Meta-analysis • Sleeve lobectomy • Pneumonectomy • Morbidity and mortality • Survival analysis • Non-small cell lung cancer


   1. Introduction
Top
 Abstract
 1. Introduction
2. Materials and methods
 3. Results
 4. Discussion
 References
 
Lung cancer, accounting for 32% and 24% of cancer deaths in men and women, respectively, represents the leading cause of cancer death worldwide. Approximately 75–80% of lung cancers are of the non-small cell lung cancer (NSCLC) histology. The overall 5-year survival rate of NSCLC is only 8–14% when diagnosed and this rate increases to 40% after complete surgical resection [1,2]. Great efforts have been made in the surgical procedure for the treatment of NSCLC. In 1933, Graham and Singer [3] performed the first successful one-stage pneumonectomy (PN) for lung cancer. After this successful operation, penumonectomy became the standard treatment for central or locally advanced NSCLC for the next two decades. Price-Thomas [4] reported in 1947 the first bronchial sleeve resection for carcinoid tumor and Allison [5] performed the first sleeve lobectomy (SL) for bronchogenic carcinoma in 1954. Since then, SL has been regarded as the standard management of NSCLC for patients who were believed to be incapable of tolerating a PN because of an inadequate pulmonary reserve, resulting from lung–parenchyma-sparing advantage.

Currently, SL is prone to be performed for such patients who could tolerate PN. Several studies have shown that long-term survival after SL is at least similar to or even better than that after PN with lower postoperative risks, better preservation of lung function, and better quality of life [6–8]. However, SL has been reported with a high bronchial postoperative complication [9–11] and a high operative mortality relative to PN [10,12,13], owing to the complex surgical procedure for SL. In contrast, other study pointed out that PN is associated with significant morbidity and mortality [14], including postpeumonectomy lung edema, acute respiratory distress syndrome (ARDS), bronchopleural fistula, and postpeumonectomy syndrome. Nonetheless, there are several reports revealing that PN does not adversely influence long-term survival [15,16].

In addition, when lung tumors involve not only the airway but also the central vascular structures, in particular the pulmonary artery, SL concomitant with pulmonary artery reconstruction (PAR) (double sleeve) remains as the only alternative to PN. However, this procedure has been slower to gain acceptance than reconstruction of bronchus alone.

This study aims to assess whether SL concomitant with or without PAR or PN offers a low morbidity and mortality and better long-term survival by mata-analysis which evaluate the published literature in a qualitative and quantitative way by comparing and integrating the results of different studies.


   2. Materials and methods
Top
 Abstract
 1. Introduction
 2. Materials and methods
3. Results
 4. Discussion
 References
 
2.1 Strategy of literature search
A PubMed search of studies published in English from 1996 to 2006 and reporting on SL for NSCLC and comparing this to PN was performed using the key words ‘sleeve resection or lobectomy and/or pneumonectomy’ and ‘lung neoplasm’. Studies on PAR were also searched by the key words ‘pulmonary artery reconstruction or pulmonary artery sleeve resection or bronchovascular sleeve resection’. All abstracts were reviewed and the ‘related articles’ function was employed to broaden the search on appropriate abstracts.

2.2 Data extraction
Two reviewers independently extracted the following data from each study: first author, year of publication, study population characteristics, study size, study design, definition of mortality and recurrences, inclusion and exclusion criteria, time of follow-up, and survival data.

Survival data were sourced from the literature. The summary statistics included the log hazard ratios (log HRs) or hazard ratio (HRs) and associated variances. Estimation of the log HR and variance were from the published survival curves according to the methods suggested by Parmar et al. [17]. Briefly, survival probabilities for SL and PN were read from the survival curves at pre-specified non-overlapping time points. The number of patients alive and at risk was calculated for each time interval, as well as the effective number of patients censored, to generate the log HR and its variance during each time interval. As the log HRs are independent, an estimate of the overall log HR for each study can be given by a weighted sum of the individual estimates of the log HR during each time interval, using the inverse-variance weighting method [18].

2.3 Inclusion and exclusion criteria
Studies included in our analysis had to:

1. Compare surgical results of SL concomitant with or without PAR versus PN techniques.
2. Be non-small cell lung cancer.
3. Report on PAR alone or associated with lobectomy or bronchial sleeve resection.
4. Report on at least one of the outcome measures mentioned below.

Studies were discarded, if:

1. Studies concerned malignancies other than non-small cell lung cancer.
2. Patients were subjected to carinal or tracheal resection.
3. Studies that displayed a zero for the outcomes of interest in both SL and PN groups.
4. Studies that did not report outcomes for PAR in the analysis of PAR.

2.4 Definition of outcomes of interest
SL and PN were compared with postoperative mortality, postoperative complications, locoregional recurrences, as well as survival outcomes. Survival outcome measures were analyzed by the difference of survival of the two techniques at 1 and 5 years, the difference of survival in patients with pN0 or pN1 and pN2 at 1 and 5 years, and overall survival.

2.5 Statistical analysis
Meta-analysis was carried out using odds ratio (OR), risk difference (RD), and HR as the summary statistics. The OR represents the odds of an adverse event occurring in the treatment (SL) group in comparison to the reference (PN) group. The RD means the difference of survival of patients subjected to two techniques at 1 and 5 years. An OR of less than 1, while a RD more than 0 favors SL group, and the point of estimate of the OR and RD is considered statistically significant at the p < 0.05 level if the 95% confidence interval does not include the value 1 or 0, respectively. Yate's correction was used for those studies without events of interest in one of the two groups, which results in problems with the calculation of ratios and their standard error of the treatment effect. To solve this problem, we have added the conventional 0.5 to each cell in the contingency table of these trials [19].

In our study both fixed and random effect models were employed. The fixed effect model is based on the assumption that the treatment effect in each study is same, whereas the random effect model is based on the assumption that there is variation between studies. Thus, the ratios calculated in the random effect model are more conservative than those calculated in the fixed effect model [20]. In a meta-analysis of surgical research, the random effect model is preferable due to a number of sources of heterogeneity (HG) including varying risk profiles, selection criteria for each surgical technique, study design, study start date, and duration of follow-up [21].

The most appropriate summary statistics for survival-type data are the log HR and its variance, which allows for both censoring and time to an event [17]. Survival data were pooled by the DerSimonian and Laird method to produce a random effect meta-analysis. Heterogeneity between studies was investigated by the standard chi-squared Q-test. Pooled survival curve [22] across all studies for SL versus PN was generated, utilizing the methodology to estimate the log HR; however, the survival probabilities were extracted at a common series of time points, then pooled.

To determine clinical benefits of two techniques, three parameters were calculated: absolute reduction (ARR), which in this case is the difference in the incidence of events between SL and PN groups; number needed to treat (NNT), which is the number of patients who must be treated (in this case to be operated by SL) in order to prevent one event (NNT = 1/ARR); and median survival times and associated variances which were estimated from the pooled survival curve.

Analysis was performed by SPSS version 13.0 for Windows (SPSS Inc., Chicago, IL, USA), Microsoft Excel 2002 (Microsoft Corporation, USA), and Review Manager Version 4.2 (The Cochrane Collaboration, Oxford).

To quantitatively assess heterogeneity between studies, three strategies were employed as reported previously [21]. First, data were re-analyzed using both fixed- and random effect models. Secondly, publication bias was assessed by funnel plot. Thirdly, subgroup analysis was used for sensitive analysis. In subgroup analysis, the following variables were used:

1. All studies
2. Matched study
3. Study size (more than 50 patients in each group)
4. Year of publication (inclusive of or greater than 2000)


   3. Results
Top
 Abstract
 1. Introduction
 2. Materials and methods
 3. Results
4. Discussion
 References
 
3.1 Selected studies
Thirteen articles [6,8,10–12,23–30] published between 1996 and 2006 met our inclusion criteria comparing SL versus PN for NSCLC. One of these was excluded, resulting from comparing SL versus sleeve pneumonectomy [27]. Finally, 12 studies, including 3 matched studies [6,10,24] and 1 prospective study [8], were suitable for meta-analysis, with a combined total of 2984 subjects, of which 876 (29%) underwent SL and 2108 (71%) underwent PN. A small portion of patients (8.9%) in six studies [6,8,10,11,12,24] in SL group underwent SL combined with PAR. Five studies [30–34] including 202 subjects were retrieved reporting on PAR alone or associated with lobectomy or bronchial sleeve resection, among which there were 164 patients (81.2%) subjected to PAR concomitant with SL.

In Tables 1 and 2 are the characteristics of these studies. The distribution of stage between the SL group and the PN group was significantly different (stages I, II, and III: 37%, 37%, and 26% for SL; 20%, 33%, and 47% for PN; p < 0.001; 10 reports). Sex ratios for the two surgical groups showed statistically significant difference as well (male/female: 65%/35% for SL; 71%/29% for PN; p < 0.001; 11 reports). However, there was no difference in mean ages for the two groups (61.6 years for SL; 60.8 years for PN), although age distributions were not available.


View this table:
[in this window]
[in a new window]

  		Table 1 Study characteristics
 

View this table:
[in this window]
[in a new window]

  		Table 2 Review of the recent literatures on pulmonary artery reconstruction
 
3.2 Postoperative mortality
The postoperative mortality was reported in 12 studies [6,8,10–12,23–26,28–30] and meta-analysis of the resultant data showed the pooled postoperative mortality in SL was 3.5% (31/876) as compared with 5.7% (121/2108) in PN, but did not reach statistical significance (OR: 0.65; 95% CI: 0.42–1.01), as shown in Fig. 1 . The calculated ARR for SL versus PN was 2.2%, indicating the NNT would be 45. Similar results were achieved when only recent studies [6,8,10–12,26,28–30] were considered (incidence of 3.5% (25/717) in SL vs 5.7% (107/1881) in PN; OR: 0.66; 95% CI: 0.41–1.08). This finding was reproduced in the case of matched studies [6,10,24] where the incidence of postoperative mortality was 2.2% (3/138) in SL versus 3.6% (5/138) in PN (OR: 0.73; 95% CI: 0.18–3.00). With regard to the studies [6,11,12,23,25,28–30] with more than 50 patients in each group, postoperative mortality was 3.0% (22/722) in SL compared with 5.7% (104/1822) in PN, suggesting that this is statistically significant (OR: 0.55; 95% CI: 0.33–0.91).


Figure 1
View larger version (24K):
[in this window]
[in a new window]

  		Fig. 1. Meta-analysis of selected studies comparing postoperative mortality, complications, and locoregional recurrences between the sleeve lobectomy group and the pneumonectomy group.

 
The weighted mean operative mortality reported in four studies [30–32,34] for PAR was 3.3%, which was not statistically significant as compared with SL group, while it was significantly lower than that of PN (p < 0.05).

3.3 Postoperative complications
Eight studies [6,10–12,23,24,26,29] reported the incidence of postoperative complications, and meta-analysis of these studies gave a pooled incidence of 31.3% (154/492) with SL and 31.6% (245/776) with PN (OR: 1.01; 95% CI: 0.70–1.44), which was not statistically significant (Fig. 1). The calculated ARR for SL versus PN was 0.3%, indicating the NNT would be 333. Subgroup analysis of studies with at least 50 patients in each group [6,11,12,23,29] produced an incidence of 30.8% (116/376) in SL as compared with 33.6% (192/571) in PN (OR: 0.93; 95% CI: 0.59–1.48), not revealing any significant difference between the two groups. Similar results were achieved when only recent studies [6,10–12,26,29] were considered (incidence of 36.3% (142/391) in SL vs 33.3% (230/691) in PN; OR: 1.08; 95% CI: 0.72–1.63). This finding was reproduced in the case of matched studies [6,10,24] where the incidence of postoperative complications was 26.8% (37/138) in SL versus 26.8% (37/138) in PN (OR: 0.89; 95% CI: 0.35–2.27).

The weighted mean postoperative complications reported in three studies [31,32,34] for PAR was 32.4%, which was neither statistically significant as compared with SL group, nor was it significantly lower than that of PN.

3.4 Locoregional recurrences
This was reported in six studies [6,10,12,24,28,29] and meta-analysis of the resultant data showed the pooled locoregional recurrence in SL was 16.1% (72/447) as compared with 27.8% (402/1443) in PN, but did not reach statistical significance (OR: 0.91; 95% CI: 0.45–1.82), as shown in Fig. 1. The calculated ARR for SL versus PN was 11.7%, indicating the NNT would be 9. Similar results were achieved when only recent studies [6,10,12,28,29] were considered (incidence of 16.5% (69/418) in SL vs 28.0% (396/1414) in PN; OR: 1.02; 95% CI: 0.46–2.26). This finding was reproduced in the case of matched studies [6,10,24] where the incidence of locoregional recurrences was 17.0% (23/135) in SL versus 11.8% (16/136) in PN (OR: 1.29; 95% CI: 0.30–5.63). Considering the studies [6,12,28,29] with more than 50 patients in each group, the incidence of locoregional recurrences was 14.5% (54/372) with SL compared with 28.7% (392/1367) with PN, suggesting that this is statistically significant (OR: 0.58; 95% CI: 0.42–0.80).

3.5 Overall survival
Survival data from 10 studies [6,10–12,24–26,28–30] were extracted, including 2708 patients, 761 in the SL group and 1947 in the PN group. A summary of individual studies and overall pooled survival is shown in Fig. 2 . The estimated combined HR for overall survival in 10 studies was 0.70 (95% CI: 0.62–0.79) in favor of SL group, which was statistically significant. Subgroup analysis of studies with at least 50 patients in each group [6,11,12,25,28–30] (HR: 0.69; 95% CI: 0.61–0.78), and only recent studies [6,10–12,26,28–30] (HR: 0.71; 95% CI: 0.63–0.80) reached statistically significant difference. This result was reproduced in the case of matched studies [6,10,24], where the HR was 0.66, with 95% CI of 0.46–0.96.


Figure 2
View larger version (11K):
[in this window]
[in a new window]

  		Fig. 2. Meta-analysis of overall survival hazard ratios in individual studies and overall.

 
The pooled survival curves (10 reports for SL and PN, 4 reports for PAR [30–32,34]) based on Kaplan–Meier calculations are shown in Fig. 3 . The median overall survival was 60 months for the SL group, 26 months for the PN group, and 30 months for the PAR group. The overall 5-year survival was 50.3% for the SL group, 30.6% for the PN group, and 38.7% for the PAR group.


Figure 3
View larger version (12K):
[in this window]
[in a new window]

  		Fig. 3. Overall survival curves.

 
3.6 Difference of survival at 1 year
Differences of survival at 1 year were extracted from 10 studies [6,10–12,24–26,28–30] and meta-analysis of these studies showed a combined RD was 0.10 (95% CI: 0.07–0.14) in favor of the SL group (Fig. 4 ). Similar results were achieved when only recent studies [6,10–12,26,28–30] were considered (RD: 0.10; 95% CI: 0.05–0.14). This finding was reproduced in the case of the studies [6,11,12,25,28–30] with more than 50 patients in each group, where the RD was 0.12 (95% CI: 0.08–0.15). Concerning the matched studies [6,10,24], the pooled RD was 0.06 (95% CI: –0.04–0.16), suggesting that this is not statistically significant.


Figure 4
View larger version (27K):
[in this window]
[in a new window]

  		Fig. 4. Meta-analysis of selected studies comparing differences of survival in patients with pN0 or pN1 and pN2 at 1 and 5 years between the sleeve lobectomy group and the pneumonectomy group.

 
3.7 Difference of survival at 5 years
Differences of survival at 5 years were extracted from nine studies [6,10–12,25,26,28–30] and meta-analysis of these studies showed a combined RD was 0.12 (95% CI: 0.07–0.18) in favor of the SL group (Fig. 4). Similar results were achieved when only recent studies [6,10–12,26,28–30] were considered (RD: 0.15; 95% CI: 0.10–0.19). This finding was reproduced in the case of the studies [6,11,12,25,28–30] with more than 50 patients in each group, where the RD was 0.12 (95% CI: 0.05–0.19). Considering the matched studies [6,10], the pooled RD was 0.12 (95% CI: –0.01–0.24), which was not statistically significant.

3.8 Differences of survival in patients with pN0 or pN1 and pN2 at 1 and 5 years
Two studies [6,11] reported the survival in patients with pN0, pN1, and pN2 at 1 and 5 years, as shown in Fig. 4. Meta-analysis of difference of survival in patients with pN0 or pN1 at 1 year demonstrated a pooled RD (SL vs PN) was 0.03 (95% CI: –0.08–0.13), which was not statistically significant. When difference of survival in patients with pN2 at 1 year was considered, the combined RD was 0.13 (95% CI: 0.00–0.25), and just reached statistical significance. Meta-analysis of difference of survival in patients with pN0 or pN1 at 5 years showed that pooled RD was 0.21 (95% CI: 0.07–0.36), which was statistically significant. Considering difference of survival in patients with pN2 at 5 years, the combined RD was 0.06 (95% CI: –0.10–0.21), indicating there was no difference between two groups.

3.9 Sensitivity analysis
In our study both fixed- and random effect models were employed for sensitivity analysis. The results for events in the SL group and PN group are shown in Table 3 . As for postoperative mortality, the heterogeneity test showed a lowest heterogeneity when only recent studies were considered (HG = 7.42, p = 0.49). Similar results were achieved when the studies of more than 50 patients in each group were concerned (HG = 5.47, p = 0.60). This finding was reproduced in the case of matched studies (HG = 1.64, p = 0.44). With regard to postoperative complications, the heterogeneity test showed HG was 5.05 (p = 0.08) for matched studies, 8.27 (p = 0.08) for the studies of more than 50 patients in each group, and 9.84 (p = 0.08) for recent studies. In the case of locoregional recurrences, the sensitivity analysis resulted in significant heterogeneity when the matched studies or the recent studies were considered, with HG of 7.24 (p = 0.02), 13.66 (p = 0.008), respectively, while still remaining relatively low heterogeneity (HG = 1.59, p = 0.66) in the studies of more than 50 patients in each group.


View this table:
[in this window]
[in a new window]

  		Table 3 Results of sensitivity analysis for events in sleeve lobectomy and pneumonectomy
 
When it comes to difference of survival at 1 year, the sensitivity analysis showed HG was 2.86 (p = 0.24) for matched studies, 5.32 (p = 0.50) for the studies of more than 50 patients in each group, and 8.51 (p = 0.29) for recent studies. This finding was repeated in the sensitivity analysis of difference of survival at 5 years, with HG of 0.11 (p = 0.74) for matched studies, HG of 12.64 (p = 0.05) for the studies of more than 50 patients in each group, and HG of 7.70 (p = 0.36) for recent studies.

Publication bias was assessed using a funnel plot (Fig. 5 ) which showed the 12 studies used in our meta-analysis. The scatter plot resembles a symmetric inverted funnel (the 95% confidence interval), in which the treatment effects estimated from individual studies on the horizontal axis (OR), against a measure of study size on the vertical funnel (SE [log OR]). The plot is on the basis that precision in the estimation of the underlying treatment effect will increase with the increment of the sample size of the component studies [21].


Figure 5
View larger version (9K):
[in this window]
[in a new window]

  		Fig. 5. Funnel plot test of the 12 studies included in our meta-analysis.

 

   4. Discussion
Top
 Abstract
 1. Introduction
 2. Materials and methods
 3. Results
 4. Discussion
References
 
Our meta-analysis study included a total of 2984 subjects, of which 876 (29%) underwent SL and 2108 (71%) underwent PN. A sample group of this size would otherwise be impossible to accumulate in a reasonable length of time in a single institute. The stage distribution between the SL group and the PN group was significantly different (stages I, II, and III: 37%, 37%, and 26% for SL; 20%, 33%, and 47% for PN; 10 reports), indicating that SL is prone to be performed in early-stage NSCLC and may have a better prognosis.

The postoperative mortality, postoperative complications, and locoregional recurrences between the two groups did not differ significantly. It is worth noting that the postoperative mortality after SL (3.5%) approached statistical significance as compared with 5.7% after PN. Furthermore, in the fixed effect model, the postoperative mortality between the two groups was significantly different, which is in line with a meta-analysis [35] of comparisons between SL and PN in stage I and II NSCLC. In that study, the weighted mean postoperative mortality after SL was 4.1%, and 6.0% after PN.

In the time-to-event analysis, the HR is one of the most appropriate effect indexes to evaluate time-related effect of surgical procedures [17]. The death risk decreased by 30% in SL group compared with PN group. Moreover, as for survival probability difference, there was an increase trend, that was 12% at 1 year and 19% at 5 years. Median overall survival was increased in the SL group as well (60 months vs 26 months). However, the previous meta-analysis study [35] showed that 5-year survival rates were similar for the two techniques (SL: 51% vs PN: 49%) when corrections for stage distribution were calculated, that use of 5-year survival rate as an outcome in the decision model slightly favored SL over PN for managing stage I and II NSCLC, and that the mean median survivals were 70.5 months for SL and 55.2 months for PN, which was statistically significant.

The relationship between survival and lymph node status is unclear after patients were subjected to SL or PN. Our study demonstrated that survival difference in patients with pN2 at 1 year just reached statistical significance, that in patients with pN0 or pN1 at 5 years showed statistical difference, and that in patients with pN2 at 5 years indicated there was no difference between two groups. Our results were in agreement with another study [28] which reported that there was significant difference in 5-year survival between SL and PN for patients with pN0 or pN1, whereas 5-year survival for patients with pN2 or stage III disease was not significant. Together with the results indicate that a more radical operation such as PN is not a more appropriate procedure, even in higher stage tumors.

When only the matched studies were concerned, there were no significant differences for postoperative mortality, postoperative complications, locoregional recurrences, survival difference at 1 year, or survival difference at 5 years between the two techniques. Therefore, further studies with strict criteria and good study design are needed to investigate the clinical benefits of two techniques.

There have been rare reports on PAR in the literature so far, which may in part due to high risk, complicated surgical procedures, and decreased long-term survival [13]. Our meta-analysis on recent results showed that the surgical outcomes of PAR were not only comparable to SL with 3.3% for the operative mortality, 32.4% for the postoperative complications, and 38.7% for the 5-year survival, but better than those of PN. These indicate that PAR combined with SL is as safe as SL for the treatment of NSCLC.

The limitations of our meta-analysis were as follows. First, definitions for postoperative mortality and locoregional recurrences were different, as illustrated in Table 1. Secondly, stage distribution between the SL group and the PN group was quite different, which might lead to unreliable results and favor the SL group. Thirdly, no randomized data were available for comparing SL and PN and the treatment protocol for NSCLC was somewhat different among institutions, all of which made the analysis and comparison difficult. Fourthly, it is of great importance to bear in mind publication bias, especially in meta-analytic research based on published retrospective studies. Fifthly, the absence of generalized and commonly shared guidelines for the set-up and data reporting also disturbed our meta-analysis. Sixthly, the population in PAR group is quit small as compared with the SL or PN group.

In the present study, 16 of the 17 studies were retrospective because a randomized prospective trial is not possible to perform. The reason for this difficulty is: (1) the final decision to perform SL or PN has to be taken during the operative procedure; (2) the small number of cases available for study; and (3) the definition of eligibility. Despite the obvious bias related to retrospective nature, meta-analysis of these non-randomized studies provides the most reliable information, which is a suggested approach to search for the ‘best evidence’ in the absence of well-controlled randomized studies [36].

In summary, our study demonstrates that SL with or without PAR is effective and can be accomplished safely in selected patients without increasing the morbidity and mortality as compared to PN, that SL even with PAR offers better long-term survival than does PN, and that a more radical operation such as PN is not a more appropriate procedure, even in higher stage tumors.


   References
Top
 Abstract
 1. Introduction
 2. Materials and methods
 3. Results
 4. Discussion
 References
 

  1. Mezzetti M, Panigalli T, Giuliani L, Raveglia F, Giudice FL, Meda S. Personal experience in lung cancer sleeve lobectomy and sleeve pneumonectomy. Ann Thorac Surg 2002;73:1736-1739.[Abstract/Free Full Text]
  2. Greenlee RT, Murray T, Bolden S, Wingo PA. Cancer statistics. CA Cancer J Clin 2000;50:7-33.[Abstract]
  3. Graham EA, Singer JJ. Successful removal of an entire lung for carcinoma of the bronchus. JAMA 1933;101:1371-1374.
  4. Price-Thomas C. Conservative resection of the bronchial tree. J R Coll Surg Edinb 1955;1:169-186.
  5. Allison PR. Course of thoracic surgery in Groningen. Ann R Coll Surg 1954;25:20-22.
  6. Okada M, Yamagishi H, Satake S, Matsuoka H, Miyamoto Y, Yoshimura M, Tsubota N. Survival related to lymph node involvement in lung cancer after sleeve lobectomy compared with pneumonectomy. J Thorac Cardiovasc Surg 2000;119:814-819.[Abstract/Free Full Text]
  7. Terzi A, Lonardoni A, Falezza G, Furlan G, Scanagatta P, Pasini F, Calabro F. Sleeve lobectomy for non-small cell lung cancer and carcinoid: results in 160 cases. Eur J Cardiothorac Surg 2002;21:888-893.[Abstract/Free Full Text]
  8. Martin-Ucar AE, Chaudhuri N, Edwards JG, Waller DA. Can pneumonectomy of non-small cell lung cancer be avoided? An audit of parenchymal sparing lung surgery. Eur J Cardiothorac Surg 2002;21:601-605.[Abstract/Free Full Text]
  9. Faber LP, Jensil RJ, Kittle CF. Results of sleeve lobectomy for bronchogenic carcinoma in 101 patients. Ann Thorac Surg 1984;37:279-285.[Abstract]
  10. Kim YT, Kang CH, Sung SW, Kim JH. Local control of disease related to lymph node involvement in non-small cell lung cancer after sleeve lobectomy compared with pneumonectomy. Ann Thorac Surg 2005;79:1153-1161.[Abstract/Free Full Text]
  11. Ludwig C, Stoelben E, Olschewski M, Hasse J. Comparasion of morbidity, 30-day mortality, and long-term survival after pneumonectomy and sleeve lobectomy for non-small cell lung carcinoma. Ann Thorac Surg 2005;79:968-973.[Abstract/Free Full Text]
  12. Takeda SI, Maeda H, Koma M, Matsubara Y, Sawabata N, Inoue M, Tokunaga T, Ohta M. Comparison of surgical results after pneumonectomy and sleeve lobectomy for non-small cell lung cancer. Trends over time and 20-year institutional experience. Eur J Cardiothorac Surg 2006;29:276-280.[Abstract/Free Full Text]
  13. Vogt-Moykopf I, Fritz TH, Meyer G, Bulzerbruck H, Daskos G. Bronchoplatic and angioplatic operation in bronchial carcicoma: long-term results of a retrospective analysis from 1973 to 1983. Int Surg 1986;71:211-220.[Medline]
  14. Bernard A, Deschamps C, Allen MS, Miller DL, Traastck VF, Jenkins GD, Pairolero PC. Pneumonectomy for malignant disease: factors affecting early morbidity and mortality. J Thorac Cardiovasc Surg 2001;121:1076-1082.[Abstract/Free Full Text]
  15. Ferguson MK, Karrison T. Does pneumonectomy for lung cancer adversely influence long-term survival?. J Thorac Cardiovasc Surg 2000;119:440-448.[Abstract/Free Full Text]
  16. Mizushima Y, Sugiyama S, Noto H, Kusajima Y, Yamashita R, Sassa K, Kobayashi M. Prognosis for patients with pneumonectomy or lesser resection for non-small cell lung cancer based on histologic cell type. Oncol Rep 1998;5:689-692.[Medline]
  17. Parmar M, Torri V, Stewart L. Extracting summary statistics to perform meta-analyses of the published literature for survival endpoints. Stat Med 1998;17:2815-2834.[CrossRef][Medline]
  18. Berlin JA, Laird NM, Sacks HS, Chalmers TC. A comparison of statistical methods for combining events rates from clinical trials. Stat Med 1989;8:141-151.[Medline]
  19. Yates F. Contingency tables involving small numbers and the chi-square test. J R Stat Soc 1934(Suppl. 1):217-235.
  20. DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials 1986;7:177-188.[CrossRef][Medline]
  21. Athanasious T, Aziz O, Al-Ruzzeh S, Philippidis P, Jones C, Purkayastha S, Casula R, Glenville B. Are wound healing disturbances and length of hospital stay reduced with minimally invasive vein harvest? A meta-analysis. Eur J Cardiothorac Surg 2004;26:1015-1026.[Abstract/Free Full Text]
  22. Le Chevalier T, Scagliotti G, Natale R, Danson S, Rosell R, Stahel R, Thomas P, Rudd RM, Vansteenkiste J, Thatcher N, Manegold C, Pujol JL, van Zandwijk N, Gridelli C, van Meerbeeck JP, Crino L, Brown A, Fitzgerald P, Aristides M, Schiller JH. Efficacy of gemcitabine plus platinum chemotherapy compared with other platinum containing regimens in advanced non-small-cell lung cancer: a meta-analysis of survival outcomes. Lung Cancer 2005;47:69-80.[CrossRef][Medline]
  23. Gaissert HA, Mathisen DJ, Moncure AC, Hilgenberg AD, Grillo HC, Wain JC. Survival and function after sleeve lobectomy for lung cancer. J Thorac Cardiovasc Surg 1996;11:948-953.
  24. Yoshino I, Tokoyama H, Yano T, Ueda T, Takai E, Mizutani K, Asoh H, Ichinose Y. Comparison of the surgical results of lobectomy with bronchoplasty and pneumonectomy for lung caner. J Surg Oncol 1997;64:32-35.[CrossRef][Medline]
  25. Suen HC, Meyers BF, Guthrie T, Pohl MS, Sundaresan S, Roper CL, Cooper JD, Patterson GA. Favorable results after sleeve lobectomy or bronchoplasty for bronchial malignancies. Ann Thorac Surg 1999;67:1557-1562.[Abstract/Free Full Text]
  26. Ghiribelli C, Voltolini L, Luzzi L, Paladini P, Campione A, Gotti G. Survival after bronchoplastic lobectomy for non-small cell lung cancer compared with pneumonectomy according to nodal status. J Cardiovasc Surg (Torino) 2002;43:103-108.[Medline]
  27. Mezzetti M, Panigalli T, Giuliani L, Giuliani L, Raveglia F, Giudice FL, Meda S. Personal experience in lung cancer sleeve lobectomy and sleeve pneumonectomy. Ann Thorac Surg 2002;73:1736-1739.[Abstract/Free Full Text]
  28. Deslauriers J, Gregoire J, Jacques LF, Piraux M, Guo G, Lacasse Y. Sleeve lobectomy versus pneumonectomy for lung cancer: a comparative analysis of survival and sites of recurrence. Ann Thorac Surg 2004;77:1152-1156.[Abstract/Free Full Text]
  29. Bagan P, Berna P, Pereira JC, Barthes FL, Foucault C, Dujon A, Riquet M. Sleeve lobectomy versus pneumonectomy: tumor characteristics and comparative analysis of feasibility and results. Ann Thorac Surg 2005;80:2046-2050.[Abstract/Free Full Text]
  30. Lausberg HF, Graeter TP, Tscholl D, Wendler O, Schafers HJ. Bronchovascular versus bronchial sleeve resection for central lung tumors. Ann Thorac Surg 2005;79(4):1147-1152.[Abstract/Free Full Text]
  31. Shrager JB, Lambright ES, McGrath CM, Wahl PM, Deeb ME, Friedberg JS, Kaiser LR. Lobectomy with tangential pulmonary artery resection without regard to pulmonary function. Ann Thorac Surg 2000;70(1):234-239.[Abstract/Free Full Text]
  32. Rendina EA, Venuta F, De Giacomo T, Ciccone AM, Moretti M, Ruvolo G, Coloni GF. Sleeve resection and prosthetic reconstruction of the pulmonary artery for lung cancer. Ann Thorac Surg 1999;68(3):995-1001.[Abstract/Free Full Text]
  33. Chunwei F, Weiji W, Xinguan Z, Qingzen N, Xiangmin J, Qingzhen Z. Evaluations of bronchoplasty and pulmonary artery reconstruction for bronchogenic carcinoma. Eur J Cardiothorac Surg 2003;23(2):209-213.[Abstract/Free Full Text]
  34. Nagayasu T, Matsumoto K, Tagawa T, Nakamura A, Yamasaki N, Nanashima A. Factors affecting survival after bronchoplasty and broncho-angioplasty for lung cancer: single institutional review of 147 patients. Eur J Cardiothorac Surg 2006;29(4):585-590.[Abstract/Free Full Text]
  35. Ferguson MK, Lehman AG. Sleeve lobectomy or pneumonectomy: optimal magnagement strategy using decision analysis techniques. Ann Thorac Surg 2003;76:1782-1788.[Abstract/Free Full Text]
  36. Rizzoli G, Schiavon L, Bellini P. Does the use of bilateral internal mammary artery (IMA) grafts provide incremental benefit relative to the use of a single IMA graft? A meta-analysis approach. Eur J Cardiothorac Surg 2002;22:781-786.[Abstract/Free Full Text]



This article has been cited by other articles:


Home page
 Eur. J. Cardiothorac. Surg.Home page
F. Rea, G. Marulli, M. Schiavon, A. Zuin, A.-M. Hamad, G. Rizzardi, E. Perissinotto, and F. Sartori
A quarter of a century experience with sleeve lobectomy for non-small cell lung cancer
Eur. J. Cardiothorac. Surg., September 1, 2008; 34(3): 488 - 492.
[Abstract] [Full Text] [PDF]

Home page
 Ann. Thorac. Surg.Home page
T. Gudbjartsson, E. Gyllstedt, A. Pikwer, and P. Jonsson
Early surgical results after pneumonectomy for non-small cell lung cancer are not affected by preoperative radiotherapy and chemotherapy.
Ann. Thorac. Surg., August 1, 2008; 86(2): 376 - 382.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to Personal Folders
Right arrow Download to citation manager
Right arrow Permission Requests
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Ma, Z.
Right arrow Articles by Cheng, H.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Ma, Z.
Right arrow Articles by Cheng, H.
Related Collections
Right arrow Lung - cancer
Right arrow Trachea and bronchi

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