P191 A meta-analysis of adjusted and unadjusted observational studies of sleeve lobectomy vs pneumonectomy for non-small cell lung cancer

    loading  Checking for direct PDF access through Ovid


Introduction and Objectives

A previous (published in 2007) meta-analysis of unadjusted results from observational studies suggested that sleeve lobectomy offered better long-term survival than did pneumonectomy for non-small cell lung cancer (NSCLC). Since the meta-analysis was conducted, a number of studies, which included ones providing adjusted mortality data, have been published to date. We performed an updated meta-analysis of sleeve lobectomy vs pneumonectomy for long-term mortality in NSCLC, combining separately adjusted and unadjusted results.


The MEDLINE and EMBASE databases and the Cochrane Library and Central Register of Controlled Trials were searched using PubMed and OVID. Studies considered for inclusion met the following criteria: the design was a study comparing sleeve lobectomy vs pneumonectomy; the study population was patients with NSCLC; and main outcomes included long-term all-cause mortality. From each individual study, hazard ratios (HRs) for mortality and 95% CIs were abstracted. Study-specific estimates were combined using inverse variance-weighted averages of logarithmic HRs.


Our search identified no randomised trials and 14 observational comparative studies that included 7 ones providing adjusted mortality data. Adjustment methods included matching (with propensity score or tumour location and invasion) and multivariate Cox proportional hazard regression. Separately pooled analysis of seven adjusted (1013 patients) and seven unadjusted studies (2278 patients) demonstrated respectively a statistically significant lower all-cause mortality by 36% and 33% associated with sleeve lobectomy relative to pneumonectomy in fixed effects models (adjusted HR, 0.64; 95% CI 0.53 to 0.77; p<0.00001; unadjusted HR, 0.67; 95% CI 0.58 to 0.77; p<0.00001; Abstract P191 figure 1). There was minimal study heterogeneity and accordingly little difference in the pooled result from random-effects modelling. When data from all the 14 studies (3291 patients) were pooled using a fixed-effects model, sleeve lobectomy was associated with lower all-cause mortality by 34% relative to pneumonectomy that remained statistically significant (HR, 0.66; 95% CI 0.59 to 0.74; p<0.00001).


Sleeve lobectomy is likely to have a benefit for long-term all-cause mortality over pneumonectomy in NSCLC. Sleeve lobectomy rather than pneumonectomy should be considered for anatomically suited NSCLC.

Related Topics

    loading  Loading Related Articles