Type of resection and prognosis in lung cancer. Experience of a multicentre study

Analysis of prognosis of the different types of resections for lung cancer defined by the Bronchogenic Carcinoma Cooperative Group of the Spanish Society of Pneumology and Thoracic Surgery (GCCB-S).

From October 1993 to September 1997, 2994 patients with bronchogenic carcinoma who underwent thoracotomy were prospectively recruited by the GCCB-S. Prior to recruitment, the GCCB-S had defined two types of non-resectional operations (diagnostic and exploratory thoracotomies) and three types of resections (complete-CR-: free resection margins, mediastinal nodal dissection, no extracapsular nodal involvement, no involvement of most distant removed nodes; relatively incomplete-RIR-: free resection margins, no mediastinal nodal dissection, unremoved nodes, involvement of most distant removed nodes, positive pleural effusion with no pleural implants; and incomplete-IR-: positive resection margins, extracapsular nodal involvement, unremoved positive nodes, positive pleural effusion with pleural implants). For survival analyses, patients with small cell carcinoma, induction therapy, postoperative mortality, unclassified operation, or lost to follow-up were excluded. The total number of evaluable patients was 2543.

In 1047 (97%) patients, RIR was defined because they had undergone a lesser nodal evaluation than mediastinal nodal dissection. Five-year survival and 95% confidence interval were: diagnostic thoracotomy 11% (0-30%), exploratory thoracotomy 5% (1-9%), IR 20% (14-26%), RIR 43% (39-47%), and CR 45% (41-49%). Differences between IR and CR or RIR were statistically significant (P<0.0001), but those between CR and RIR were not (P=0.18).

CR and RIR should be combined in a single category as complete resection, because they do not discriminate prognostic differences.